Technical Field
[0001] Embodiments described herein generally relate to determining presence of a gas. More
particularly, although not exclusively, embodiments relate to obtaining polarisation-dependent
radiation intensity measurements from which the presence of a gas may be determined.
Some embodiments relate to remotely determining presence of a gas. Some embodiments
relate to determining the concentration of the gas. The present invention relates
to an apparatus and a method for determining presence of a gas in a gas detection
volume.
Background
[0002] Remote sensing of gases in the Earth's atmosphere is essential for air pollution
monitoring, source detection and gas leak detection. Many optical gas sensing instruments
are based on spectroscopic measurements in which trace gases such as NO
2 or CH
4, can be detected through their specific spectral features. In order to achieve a
two-dimensional image of the presence of trace gas it is common for a spectrometer
to use a scanning mechanism in which the field of view can be scanned in both a spatial
and a spectral domain. A scanning mechanism may require substantial time for the 2D
image to be built up. Other mechanisms for obtaining a 2D image of trace gas concentration
include hyperspectral imaging and compressive sensing techniques. Such methods may
implement a dispersive or wavelength-selective element to obtain an inherent spectral
resolution sufficient to discern the spectral features of the gas.
[0003] Multi Axis Differential Optical Absorption Spectroscopy (MAX-DOAS) instruments can
be used to determine concentrations of gases in the atmosphere by utilising scattered
sunlight from multiple viewing directions. The scattered sunlight may be detected
using spectrographs which may collect light through an optical fibre and scan the
sky to record complete spectral features in all directions. Instruments which use
optical fibre may have several viewing directions at the same time. In scanning systems,
the entire system (including optical fibres) may be pointed towards the target area
to prevent transmission differences from causing erroneous readings.
[0004] A further example of gas detection involves using a multiple optical path intensity
based interferometer method such as a using a Michelson Interferometer or a Fourier
Transform Spectrometer, where an incoming beam of radiation is split into two physically
separate independent arms before being recombined. A yet further example involves
the use of a Mach-Zender interferometer which utilises different optical paths and
has two outputs which are out of phase with each other by 180°. Due to the complex
optical system associated with such interferometers, they are prone to become misaligned
and are particularly sensitive to vibrations. The multiple optical paths in such interferometers
may mean that small misalignments or movements may lead to erroneous measurements.
Vibrations and temperature changes can affect different optical paths differently.
As a result, such interferometric techniques may be considered unsuitable in many
environments, such as for an example in a drone, satellite or any moving vehicle,
when a large amount of stabilisation may be required. Furthermore, such interferometers
with multiple physically separate optical paths are highly sensitive to temperature
and vibration changes, or require regular scanning of the interferometer path length,
resulting in temperature and vibration stabilisation being required. The stabilisation
requirements and sensitive alignments lead to a bulky set-up.
[0005] There is thus a need for a compact and robust imaging apparatus for detecting gases.
[0006] EP 3 076 160 A1 discloses an aerosol detector system for spatially resolved detection of an aerosol
distribution in an area.
[0007] WO 2016/076724 A2 discloses a method of spatially resolved detection of a gas substance in an area.
Summary
[0008] Irrespective of what is mentioned in the following description, the extent of protection
conferred is solely determined by the claims.
[0009] A first aspect of the disclosure provides an apparatus for determining presence of
a gas in a gas detection volume as defined by appended independent claim 1. Particular
embodiments of the present invention are defined by the appended dependent claims.
[0010] It may be that the apparatus further comprises a filter to selectively transmit received
radiation having a frequency within the selected frequency range. Typically the gas
has a plurality of spectral features offset in frequency from each other within the
selected frequency range. For example, an absorption or emission spectrum of the gas
may have a plurality of absorption or emission lines offset in frequency from each
other within the selected frequency range.
[0011] A second aspect of the disclosure provides a method for determining presence of a
gas in a detection volume as defined in appended independent claim 15.
[0012] It may be that the method further comprises selectively transmitting received radiation
having a frequency within the selected frequency range. Typically the gas has a plurality
of spectral features offset in frequency from each other within the selected frequency
range.
[0013] For example, an absorption or emission spectrum of the gas may have a plurality of
absorption or emission lines offset in frequency from each other within the selected
frequency range.
[0014] By providing one or more retarders to spectrally modulate polarisation of received
radiation in accordance with a plurality of polarised spectral modulation profiles
which are offset in phase from each other, and selectively (and typically separately)
detecting polarised radiation conforming to each of at least first and second of the
said polarised spectral modulation profiles, a compact arrangement can be provided
which enables polarisation-dependent radiation intensity measurements to be performed
from which the presence of the gas in a gas detection volume can be determined. Such
an apparatus can also be provided with a tolerable sensitivity to ambient temperature
changes without having to stabilise the temperature of the apparatus itself. By providing
apparatus with tolerable sensitivity to ambient temperature changes, any calibration
required for the apparatus is more robust and does not need to be repeated for every
change in ambient temperature. The apparatus is also more tolerant to vibration. As
the polarisation of received radiation is spectrally modulated in accordance with
the polarised spectral modulation profiles by the one or more retarders, two polarised
components of the same beam of the received radiation travel along essentially the
same trajectory with a path length difference. Intensity modulation of the beam of
received radiation (typically in accordance with a said polarised spectral modulation
profile) may then be obtained by a polariser filtering the radiation output by the
one or more retarders for a respective polarisation state. Accordingly, an interferometer
with multiple optical paths is not required in order to provide the radiation with
the said plurality of polarised spectral modulation profiles, which helps the apparatus
to have more tolerable temperature and vibration sensitivity. By contrast, splitting
the beam into two arms and recombining using a classical interferometric method such
as a Michelson interferometer, or using a Mach-Zender interferometer having more than
one optical path, provides a larger scope for errors due to properties of the multiple
optical paths varying differently with temperature and vibrations.
[0015] The radiation output from the one or more retarders comprises radiation having polarisation
spectrally modulated in accordance with the said plurality of polarised spectral modulation
profiles in a common beam of radiation. Typically the radiation filtered for a respective
polarisation state and conforming to the first said polarised spectral modulation
profile is filtered from a beam of radiation comprising radiation having polarisation
spectrally modulated in accordance with the at least first and second said polarised
spectral modulation profiles (and typically similarly for the radiation filtered for
a respective polarisation state and conforming to the second polarised spectral modulation
profile). By providing radiation having polarisation spectrally modulated in accordance
with the said plurality of polarised spectral modulation profiles in a common beam
of radiation, the radiation modulated in accordance with each of the polarised spectral
modulation profiles follows substantially the same optical path from the one or more
retarders to the detectors. This helps to keep the apparatus compact and avoids the
need for multiple discrete optical paths, thereby helping to improve the tolerance
of the apparatus to temperature changes and vibration.
[0016] By the radiation detectors selectively and separately detecting on different detectors
at the same time polarised radiation conforming to each of the at least first and
second of the said polarised spectral modulation profiles, the number of moving parts
of the apparatus can be reduced, making the apparatus more robust and reliable. In
addition, more sensitive and accurate gas detection can be achieved at least because
the at least first and second polarisation dependent radiation intensity measurements
are made under the same temperature, vibration, optical alignment and gas concentration
conditions.
[0017] It may be that the one or more retarders is to spectrally modulate polarisation of
received radiation in accordance with the plurality of polarised spectral modulation
profiles by causing frequency dependent interference between orthogonal polarisation
components of received radiation.
[0018] It will be understood that, when it is mentioned that radiation may be detected which
conforms to at least first and second of the said polarised spectral modulation profiles
on different detectors at the same time, the different detectors may be different
in the sense that they are discrete from each other, but they may otherwise be substantially
the same type of detector. The different detectors may be different pixels on the
same array of radiation detectors, such as a two-dimensional array of radiation detectors.
[0019] It may be that the polarised spectral modulation profiles are linearly polarised
spectral modulation profiles. It may be that the one or more radiation detectors detect
radiation output from the one or more retarders filtered for a respective linear polarisation
state. It may be that the one or more detectors selectively (and typically separately,
for example on different detectors at the same time) detect linearly polarised radiation
conforming to each of at least first and second of the said polarised spectral modulation
profiles to thereby provide at least respective first and second polarisation-dependent
radiation intensity measurements from which the presence of the gas can be determined.
[0020] It may be that the one or more polarisers comprise one or more polarising beam splitters
which each provide at separate outputs polarised radiation at different predefined
(e.g. orthogonal) linear polarisations, for example to respective detectors. It may
be that the one or more polarisers comprise a plurality of linear polarising filters
which each have a single output to provide radiation having a single predefined linear
polarisation, for example to a respective detector.
[0021] Typically, when polarisation of radiation is spectrally modulated by the one or more
retarders in accordance with each of a plurality of polarised spectral modulation
profiles, the polarisation of radiation is spectrally modulated such that, when the
spectrally modulated radiation is filtered at a polarisation angle corresponding to
a respective said polarised spectral modulation profile, the intensity of the filtered
radiation varies (typically substantially periodically) with frequency of radiation
in accordance with that profile. Typically each of the polarised spectral modulation
profiles, and thus the intensity of the radiation filtered at a polarisation angle
corresponding to that polarised spectral modulation profile, comprises alternating
maxima and minima which are substantially periodic with frequency of radiation, typically
within the selected frequency range. Typically the maxima and minima of the first
said polarised spectral modulation profile are offset in phase from the maxima and
minima of the second said polarised spectral modulation profile. It may be that the
maxima and minima are interference maxima and minima caused by frequency dependent
interference between orthogonal polarisation components caused by the one or more
retarders.
[0022] The said polarised spectral modulation profiles have alternating maxima and minima
which are substantially periodic with frequency of radiation, within the selected
frequency range, the frequency spacing between successive maxima being substantially
equal to the frequency spacing between peaks of a pair of spectral features of the
gas in the gas detection volume within the selected frequency range. It may be that
the first polarised spectral modulation profile has successive minima the frequency
spacing between which is substantially equal to the frequency spacing between peaks
of a pair of spectral features of the gas within the selected frequency range. It
may be that the respective minima of the first polarised spectral modulation profile
substantially do not overlap with the peaks of the respective spectral features of
the gas in the selected frequency range.
[0023] It may be that the spectral features are absorption lines or emission lines of the
gas.
[0024] It may be that a plurality of spectral features of the gas within the selected frequency
range are at least quasi-periodic or periodic. It may be that the maxima and minima
of the polarised spectral modulation profiles are provided with a frequency spacing
which is substantially equal to the frequency spacing between peaks of spectral features
of the said plurality of spectral features within the selected frequency range. It
may be that the first polarised spectral modulation profile has successive maxima
the frequency spacing between which is substantially equal to the frequency spacing
between peaks of a pair of spectral features of the gas within the selected frequency
range. It may be that a plurality of the respective maxima of the first polarised
spectral modulation profile substantially overlap with the respective spectral features
(preferably the peaks of the spectral features) of the gas within the selected frequency
range corresponding to the said (e.g. absorption) peaks. The closer the period between
maxima (and the period between minima) of the polarised spectral modulation profiles
to the period between peaks of spectral features of the gas, the more sensitive the
resulting fractional linear polarisation to the presence of the gas.
[0025] It may be that the first polarised spectral modulation profile is substantially 180°
out of phase with the second polarised spectral modulation profile. It may be that
the second polarised spectral modulation profile has successive maxima the frequency
spacing between which is substantially equal to the frequency spacing between peaks
of a pair of spectral features of the gas within the selected frequency range. It
may be that the respective maxima of the second polarised spectral modulation profile
substantially do not overlap with the respective peaks of the spectral features of
the gas within the selected frequency range.
[0026] It may be that the one or more retarders comprise one or more retarder elements and
a quarter wave plate. It may be that the one or more retarders comprise one or more
retarder elements to provide received radiation with components which are linearly
polarised at linear polarisation angles of 0° and 90° and components which are right
and left hand circularly polarised. It may be that the one or more retarders comprise
a quarter wave plate. It may be that the quarter wave plate is to convert the components
of the radiation output by the retarder elements having linear polarisation angles
of 0° and 90° and right and left hand circularly polarisation to linearly polarised
components at linear polarisation angles of 0°, 45°, 90° and 135°. It may be that
the one or more retarder elements and quarter waveplate (where provided) together
provide the received radiation with components which are linearly polarised at linear
polarisation angles of 0°, 45°, 90° and 135°. These linear polarisation angles may
relate to modulation phases of 0°, 90°, 180° and 270°.
[0027] It may be that the plurality of polarised spectral modulation profiles comprises
at least three polarised spectral modulation profiles offset in phase from each other.
[0028] It may be that the plurality of polarised spectral modulation profiles comprises
at least four polarised spectral modulation profiles offset in phase from each other.
[0029] It may be that the one or more detectors selectively (and typically separately, for
example on different detectors at the same time) detect (or it may be that the method
comprises selectively detecting, for example on different detectors at the same time)
polarised radiation conforming to each of at least first, second and third of the
said polarised spectral modulation profiles to thereby provide at least respective
first, second and third polarisation-dependent radiation intensity measurements from
which the presence of the gas can be determined.
[0030] It may be that the radiation detectors comprise at least first, second and third
radiation detectors to selectively, separately and simultaneously detect polarised
radiation filtered for respective polarisation states, the polarised radiation detected
by each of the at least first, second and third radiation detectors conforming to
a different one of the said polarised spectral modulation profiles, to thereby provide
respective polarisation-dependent radiation intensity measurements from which the
presence of the gas can be determined.
[0031] The fractional linear polarisation of the detected radiation, and thus the presence
of the gas, can be determined from the respective at least first, second and third
polarisation-dependent radiation intensity measurements to thereby determine presence
(and typically absorption or emission strength or concentration) of the gas without
any of the first, second and third of the said polarised spectral modulation profiles
having to be aligned with the absorption lines of the gas. This helps to further reduce
the temperature sensitivity of the apparatus as temperature dependent spectral shifts
of the polarised spectral modulation profiles can be more easily tolerated.
[0032] It will be understood that the more distributed the relative phases of the first,
second and third of the said polarised spectral modulation profiles throughout a 360°
cycle, the more efficiently the fractional linear polarisation can be determined based
on the at least first, second and third polarisation-dependent radiation intensity
measurements. Accordingly, preferably respective ones of the first, second and third
polarisation-dependent spectral modulation profiles which are adjacent to each other
in phase are offset in phase from each other by substantially 120°. Preferably, the
fractional linear polarisation determined by the controller is at least 5% of the
average (e.g. mean) relative absorption depth of the spectral features of the gas
within the selected frequency range.
[0033] It may be that each of the at least first, second and third polarised spectral modulation
profiles have a plurality of maxima in the selected frequency range which do not align
(or overlap) with peaks of spectral features of the gas in the selected frequency
range.
[0034] It may be that the one or more detectors selectively (and typically separately, for
example on different detectors at the same time) detect (or it may be that the method
comprises one or more detectors selectively detecting, for example on different detectors
at the same time) polarised radiation conforming to each of at least first, second,
third and fourth of the said polarised spectral modulation profiles to thereby provide
at least respective first, second, third and fourth polarisation-dependent radiation
intensity measurements from which the presence of the gas can be determined.
[0035] It may be that the radiation detectors comprise at least first, second, third and
fourth radiation detectors to selectively, separately and simultaneously detect polarised
radiation filtered for respective polarisation states, the polarised radiation detected
by each of the at least first, second, third and fourth radiation detectors conforming
to a different one of the said polarised spectral modulation profiles, to thereby
provide respective polarisation-dependent radiation intensity measurements from which
the presence of the gas can be determined.
[0036] It may be that respective ones of the first, second, third and fourth polarisation-dependent
spectral modulation profiles which are adjacent to each other in phase are offset
in phase from each other by substantially 90°.
[0037] It may be that each of the at least first, second, third and fourth polarised spectral
modulation profiles have a plurality of maxima in the selected frequency range which
do not align (or overlap) with peaks of spectral features of the gas in the selected
frequency range.
[0038] It may be that the one or more detectors selectively (and typically separately) detect
(or it may be that the method comprises one or more detectors selectively detecting)
polarised radiation conforming to each of at n said polarised spectral modulation
profiles to thereby provide at least respective n polarisation-dependent radiation
intensity measurements from which the presence of the gas can be determined.
[0039] It may be that respective ones of the n polarisation-dependent spectral modulation
profiles which are adjacent to each other in phase are offset in phase from each other
by substantially 360/n°.
[0040] It may be that n is, for example, an integer greater than or equal to two, greater
than equal to three, greater than equal to four or greater than equal to eight.
[0041] It may be that the respective maxima of the polarised spectral modulation profiles
to which the radiation selectively detected by the one or more detectors conform substantially
do not overlap with the respective peaks of the spectral features of the gas in the
selected frequency range.
[0042] It may be that the one or more retarders are to spectrally modulate the linear polarisation
state of the received radiation to provide the received radiation with a plurality
of polarised spectral modulation profiles which are offset in phase from each other.
It may be that each of the said plurality of polarised spectral modulation profiles
is associated with a different linear polarisation angle (typically such that when
the spectrally modulated radiation is filtered at the linear polarisation angle associated
with a respective said polarised spectral modulation profile, the intensity of the
filtered radiation varies (typically substantially periodically) with the frequency
of radiation in accordance with that profile).
[0043] It may be that radiation incident on the one or more retarders at different angles
of incidence is modulated by the one or more retarders (typically one or more birefringent
elements) in accordance with respective different polarised spectral modulation profiles
which are offset in phase from each other (or it may be that the method comprises
the one or more retarders modulating radiation incident on the one or more retarders
at different angles of incidence in accordance with respective different polarised
spectral modulation profiles which are offset in phase from each other), the said
different polarised spectral modulation profiles being associated with the same (typically
linear) polarisation angle.
[0044] It may be that the one or more retarders are to spectrally modulate the linear polarisation
state of the received radiation in accordance with different polarised spectral modulation
profiles for different angles of incidence of the radiation on the one or more retarders.
[0045] It may be that the one or more retarders are to apply a combination of modulations,
such as spectral modulations, to the received radiation, such as to the polarisation
of the received radiation, in different modulation domains. For example it may be
that the one or more retarders are to apply any combination of the following modulations
to received radiation: to spectrally modulate the linear polarisation state of the
received radiation to provide the received radiation with a plurality of polarised
spectral modulation profiles which are offset in phase from each other, each of which
is associated with a different linear polarisation angle; to spectrally modulate the
linear polarisation state of the received radiation in accordance with different polarised
spectral modulation profiles for different angles of incidence of the radiation on
the one or more retarders (e.g. for radiation of the same linear polarisation angle);
to modulate the polarisation state of radiation output by the one or more retarders
(e.g. periodically) over time.
[0046] It may be that the apparatus is to apply a plurality of polarisation modulations
to the received radiation, the plurality of polarisation modulations comprising the
one or more retarders spectrally modulating the linear polarisation state of the received
radiation to provide the received radiation with a plurality of polarised spectral
modulation profiles which are offset in phase from each other, each of which is associated
with a different linear polarisation angle, and any one or more of: the one or more
retarders spectrally modulating the linear polarisation state of the received radiation
in accordance with different polarised spectral modulation profiles for different
angles of incidence of the radiation on the one or more retarders (e.g. for radiation
of the same linear polarisation angle); the apparatus modulating the polarisation
state of radiation output by the one or more retarders differently (e.g. periodically)
over time; the apparatus spatially modulating the received radiation by a plurality
of groups of radiation detectors detecting radiation output from the one or more retarders,
each of the groups of detectors comprising radiation detectors to detect radiation
output from the one or more retarders filtered for respective polarisation states
by one or more polarisers, the radiation detectors selectively and separately detecting
on different detectors at the same time polarised radiation conforming to each of
at least first and second of the said polarised spectral modulation profiles to thereby
provide at least respective first and second polarisation-dependent radiation intensity
measurements from which the presence of the gas can be determined, the respective
groups of radiation detectors being offset from each other, for example in a common
focal plane. It may be that the respective groups of radiation detectors are offset
from each other in a periodic tessellated arrangement, the periodicity of the tessellated
arrangement being in respect of the linear polarisation angles of radiation which
the detectors of the groups are configured to detect.
[0047] It may be that the spectral polarisation modulation provided by the one or more retarders
provides the received radiation with the polarised spectral modulation profiles which
allow the presence of gas to be determined, while the additional angular, temporal
and/or spatial modulation(s) helps to improve the accuracy with which radiation conforming
to those profiles can be measured.
[0048] It may be that the radiation detectors and the one or more polarisers are provided
by a pixelated polarisation camera. It may be that the pixelated polarisation camera
is to perform the said spatial modulation.
[0049] Although gas detection can be performed by way of modulation in a single modulation
domain (for example the spectral modulation applied by the one or more retarders to
provide the radiation with polarised spectral modulation profiles offset in phase
with each other, the different polarised spectral modulation profiles being associated
with different linear polarisation angles), by adding modulations in different modulation
domains to the radiation, systematic errors, for example in a fractional linear polarisation
measurement or in the polarisation-dependent intensity measurements from which the
presence of the gas is determined can be reduced or removed. This enables more accurate,
and more sensitive, polarisation measurements and thus gas detection.
[0050] It may be that the detected radiation conforming to at least one of the said at least
first and second polarised spectral modulation profiles is radiation filtered for
a first linear polarisation angle and wherein the detected radiation conforming to
at least one of the said at least first and second polarised spectral modulation profiles
is radiation filtered for a second linear polarisation angle different from the first
linear polarisation angle.
[0051] It may be that the second linear polarisation angle is offset from the first linear
polarisation angle by substantially 90° (e.g. between 80° and 100°, or between 85°
and 95° or between 88° and 92° or between 89° and 91°). This provides a substantially
180° phase difference between the polarised spectral modulation profiles associated
with the first and second linear polarisation angles, thereby enabling the maxima
of one of the polarised spectral modulation profiles to be aligned with peaks of the
spectral features of the gas while the maxima of the other polarised spectral modulation
profile are aligned with portions of the spectrum of the gas off the said peaks.
[0052] It may be that detected radiation conforming to at least one of the at least first
and second polarised spectral modulation profiles is radiation polarised at a first
linear polarisation angle which was incident on a first of the said one or more retarders
at a first angle of incidence, and wherein the detected radiation conforming to at
least one of the at least first and second polarised spectral modulation profiles
is radiation polarised at the said first linear polarisation angle which was incident
on the said first of the said one or more retarders at a second angle of incidence
different from the first angle of incidence.
[0053] It may be that the apparatus comprises one or more radiation detectors to selectively
detect radiation output from the one or more retarders filtered for a respective polarisation
state by one or more polarisers, the one or more detectors selectively detecting polarised
radiation conforming to each of a plurality of the said polarised spectral modulation
profiles to thereby provide a respective plurality of polarisation-dependent radiation
intensity measurements from which the presence of the gas can be determined. Typically
the plurality of polarised spectral modulation profiles are offset in phase from each
other. It may be that the said plurality of polarised spectral modulation profiles
comprise a plurality of polarised spectral modulation profiles associated with each
of a plurality of different polarisations (typically a plurality of linear polarisations),
the said plurality of polarised spectral modulation profiles associated with each
of the plurality of different polarisations being associated with radiation incident
on the one or more retarders at different angles of incidence.
[0054] It may be that the one or more retarders comprise a temporal polarisation modulator
(e.g. rotating half waveplate or switching liquid crystal) to (typically cyclically)
vary the (typically linear) polarisation state (or it may be that the method comprises
a temporal polarisation modulator (typically cyclically) varying the (typically linear)
polarisation state) of the radiation output from the one or more retarders with respect
to time. It may be that the temporal polarisation modulator sequentially allows linearly
polarised radiation at different angles of linear polarisation to pass through the
combination of the temporal polarisation modulator and the one or more (typically
fixed angle linear) polarisers onto the said one or more radiation detectors. Thus,
the one or more radiation detectors detecting radiation filtered for a particular
polarisation state can detect radiation in accordance with different said polarised
spectral modulation profiles over time. It may be that the temporal modulation modulator
is provided in addition to one or more retarders that perform spectral modulation
of the received radiation, the spectral modulation providing the received radiation
with a plurality of polarised spectral modulation profiles offset in phase with each
other, each of the said polarised spectral modulation profiles being associated with
a different linear polarisation angle.
[0055] It may be that the plurality of radiation detectors include at least one respective
radiation detector for (or it may be that the method comprises a plurality of radiation
detectors including at least one respective radiation detector) detecting radiation
conforming to each of the respective at least first and second polarised spectral
modulation profiles. It may be that the said plurality of radiation detectors are
provided in a tessellated arrangement.
[0056] It may be that the apparatus comprises a plurality of (typically fixed angle) linear
polarisers, each to receive (or it may be that the method comprises a plurality of
(typically fixed angle) linear polarisers each receiving) radiation having spectrally
modulated polarisation from the one or more retarders and output (or outputting) linearly
polarised radiation (e.g. at a respective linear polarisation angle) to a respective
radiation detector of the said plurality of radiation detectors, typically to provide
respective polarisation dependent intensity measurements from which presence of the
gas can be determined. It may be that the plurality of linear polarisers comprise
linear polarisers of different linear polarisation angles. It may be that the linear
polarisers of the said plurality are provided on a common focal plane.
[0057] It may be that the linear polarisers of the said plurality are offset from each other,
for example in the common focal plane.
[0058] It may be that the plurality of linear polarisers are provided in a tessellated arrangement.
[0059] It may be that the one or more polarisers are provided over corresponding detectors.
[0060] It may be that the linear polarisers of the plurality of linear polarisers are provided
over the corresponding detectors.
[0061] It may be that the said plurality of radiation detectors are provided on a common
focal plane. It may be that the radiation detectors of the said plurality are offset
from each other, for example in the common focal plane.
[0062] It may be that the apparatus comprises an adjustable (e.g. rotating) linear polariser
(i.e. adjustable to thereby adjust the linear polarisation angle of the linear polariser),
or a combination of a temporal polarisation modulator and a fixed angle linear polariser
together, to receive (or it may be that the method comprises an adjustable (e.g. rotating)
linear polariser receiving) radiation having spectrally modulated polarisation from
the one or more retarders to sequentially allow (or sequentially allowing) linearly
polarised radiation at different angles of linear polarisation to pass through the
adjustable linear polariser, or through the said combination of the temporal polarisation
modulator and fixed angle linear polariser, onto the said one or more radiation detectors.
Thus, the one or more radiation detectors can detect radiation in accordance with
different said polarised spectral modulation profiles over time.
[0063] It may be that the one or more retarders are arranged in an athermal combination.
For example, it may be that the one or more retarders comprise a plurality of retarder
(e.g. birefringent) elements at least two of which have different retardances (e.g.
birefringences), the said plurality of retarder (e.g. birefringent) elements being
combined such that temperature dependent variations in retardance (e.g. birefringence)
of the said plurality of retarder (e.g. birefringent) elements at least substantially
cancel out. It may be that the said at least two retarder (e.g. birefringent) elements
have retardances which have different relative temperature-dependences (as determined
by the coefficient of thermal expansion and the thermo-optic coefficients for n
e and n
o). It may be that the at least two retarder elements comprise at least two birefringent
elements each comprising fast and slow axes, and it may be that the fast axis of a
first said birefringent element is aligned with the slow axis of a second birefringent
element. It may be that the birefringence of the first said birefringent element is
different from the birefringence of the second birefringent element. It may be that
the change with temperature of the path length difference between the fast and slow
axes of the first said birefringent element is substantially equal to the change with
temperature of the path length difference between the fast and slow axes of the second
said birefringent element.
[0064] By arranging the one or more retarders in an athermal combination, it can be ensured
that the alignment between the maxima of the first polarised spectral modulation profile
and the peaks of spectral features of the gas (and the alignment between the maxima
of the second polarised spectral modulation profile with portions of the spectrum
of the gas off the said peaks) remains stable with changes in temperature of the apparatus.
This helps to enable the presence of gas to be determined from two polarisation-dependent
intensity measurements. When more than two polarisation-dependent intensity measurements
are provided, such alignment is not necessary. However, the athermal combination can
still help to improve the accuracy of measurements, for example by reducing or eliminating
measurement errors caused by temperature fluctuations causing corresponding fluctuations
in retardances of the one or more retarders.
[0065] It may be that the filter comprises a bandpass filter. It may be that the filter
comprises an interference filter.
[0066] It may be that the filter comprises one or more birefringent filters. It may be that
the filter comprises one or more retarder (e.g. birefringent) elements and one or
more polarisers. It may be that the filter comprises one or more retarders in an athermal
combination. For example, as above, it may be that the one or more retarders of the
filter comprise a plurality of retarder (e.g. birefringent) elements at least two
of which have different retardances (e.g. birefringences), the said plurality of retarder
(e.g. birefringent) elements being combined such that temperature dependent variations
in retardance (e.g. birefringence) of the said plurality of retarder (e.g. birefringent)
elements at least substantially cancel out.
[0067] It may be that the filter comprises a bandpass (e.g. interference) filter and one
or more birefringent filters. It may be that a transmission response of the one or
more birefringent filters has transmission minima at the lower and upper cut-off frequencies
of the bandpass filter. It may be that the one or more birefringent filters shape
the combined transmission profile of the bandpass filter and the one or more birefringent
filters between the said transmission minima.
[0068] It may be that the one or more retarders comprise one or more birefringent elements,
wherein the one or more birefringent filters comprise one or more birefringent elements,
and wherein the overall temperature dependence of the retardance of the birefringent
elements of the one or more retarders is substantially matched to the overall temperature
dependence of the retardance of the birefringent elements of the one or more birefringent
filters. It may be that the temperature dependence of the one or more retarders and/or
the temperature dependence of the one or more birefringent elements of the one or
more birefringent filter(s) match the temperature dependence of the bandpass (e.g.
interference) filter.
[0069] It may be that the radiation received by the apparatus is (e.g. linearly) polarised,
or it may be that the radiation received by the apparatus is unpolarised or it may
be that the radiation received by the apparatus is partially polarised. It may be
that the apparatus further comprises a linear polariser to receive radiation and output
linearly polarised radiation to the one or more retarders.
[0070] It may be that the apparatus further comprises a controller configured to determine
presence of the gas based on the at least first and second polarisation-dependent
intensity measurements. For example, it may be that the controller is configured to
compare the first polarisation-dependent intensity measurement to the second polarisation-dependent
intensity measurement to thereby determine the presence of the gas. It may be that
the controller is configured to determine a gas detection measurement based on the
at least first and second polarisation-dependent intensity measurements.
[0071] It will be understood that the controller typically comprises a hardware processor.
Typically the controller comprises data acquisition circuitry in data communication
with processing circuitry, or a hardware computer processor executing computer program
code (which is typically provided in a computer memory in data communication with
the hardware processor) causing the hardware computer processor, to receive and process
the polarisation-dependent intensity measurements in accordance with the functionality
of the controller as set out herein.
[0072] It may be that the controller is to determine (or it may be that the method comprises
determining) fractional linear polarisation data relating to the fractional linear
polarisation of the radiation output by the one or more retarders, typically based
on the polarisation-dependent intensity measurements. For example, it may be that
the controller is to determine (or it may be that the method comprises determining)
fractional polarisation data relating to a fractional linear polarisation of the radiation
output by the one or more retarders based on the polarisation-dependent intensity
measurements by: determining Q
meas, U
meas and I
meas based on the polarisation-dependent intensity measurements; determining Q
meas/I
meas and U
meas/I
meas from the determined Q
meas, U
meas and I
meas; applying a transformation to the [Q
meas, U
meas] co-ordinate system to provide a transformed co-ordinate system [Q
meas', U
meas'] whereby Q
meas'/I
meas>0 and U
meas'/I
meas=0; and determining the fractional linear polarisation data based on Q
meas'/I
meas,.
where:
Qmeas is the measured polarisation at 0°/90° of a Stokes vector derived from the polarisation-dependent
radiation intensity measurements.
Umeas is the measured polarisation at 45°/135° of a Stokes vector derived from the polarisation-dependent
radiation intensity measurements.
Imeas is the measured intensity of a Stokes vector derived from the polarisation-dependent
radiation intensity measurements.
[0073] It may be that applying the transformation comprises taking into account a zero target
gas concentration reference. It may be that the zero target gas concentration reference
relates to a measure of the fractional linear polarisation of the radiation output
by the one or more retarders for a zero target gas concentration in the target gas
measurement volume. It may be that the zero target gas concentration reference is
a temperature dependent zero target gas concentration reference. For example, the
zero target gas concentration reference may be dependent on the temperature of the
apparatus when the polarisation-dependent intensity measurements were made. It may
be that the method comprises measuring the temperature of the apparatus (e.g. or the
temperature of the one or more retarders), selecting a zero target gas concentration
reference in dependence on the measured temperature, and determining the fractional
linear polarisation data taking into account the selected zero target gas concentration
reference.
[0074] Thus, it may be that the fractional linear polarisation data comprises a measure
of the fractional linear polarisation of the radiation output by the one or more retarders.
It may be that the measure of the fractional linear polarisation of the radiation
output by the one or more retarders is with respect to a zero target gas concentration
reference which may be dependent on the temperature of the apparatus (in particular,
the temperature of the one or more retarders) when the polarisation dependent radiation
intensity measurements were made.
[0075] It may be that the apparatus is designed to provide a negligible U
meas (e.g. by not providing polarised radiation at linear polarisation angles of 45° and
135°). In this case, it may be that the fractional linear polarisation can be determined
from Q
meas/I
meas.
[0076] It may be that the controller is configured to determine the presence of the gas
based on the fractional linear polarisation data.
[0077] It may be that the controller is configured to determine (or it may be that the method
comprises determining) an absorption or emission strength or concentration (e.g. a
column integrated concentration in an atmospheric volume) of the gas based on the
said fractional linear polarisation data. It may be that the controller is configured
to determine (or it may be that the method comprises determining) an absorption or
emission strength or concentration of the gas (e.g. a column integrated concentration
in an atmospheric volume) based on the determined fractional linear polarisation data,
typically with reference to one or more calibration references.
[0078] It may be that the fractional linear polarisation data determined by the controller
(or by the method) is indicative of a fractional linear polarisation of at least 5%
of the average relative depth of the spectral features of the gas within the selected
frequency range.
[0079] It may be that the apparatus further comprises an objective lens for (or it may be
that the method comprises an objective lens) receiving received radiation and directing
it towards the filter.
[0080] It may be that the objective lens is a telecentric lens. In this case, radiation
received and output by the telecentric objective lens is typically incident on the
one or more retarders at an angle of incidence independent of its angle of incidence
on the telecentric objective lens.
[0081] It may be that the objective lens is a non-telecentric lens. In this case, radiation
received and output by the non-telecentric objective lens is typically incident on
the one or more retarders at an angle of incidence which is dependent on its angle
of incidence on the non-telecentric objective lens.
[0082] It may be that the apparatus comprises a plurality of groups of radiation detectors,
each of the said groups comprising one or more radiation detectors to detect (or it
may be that the method comprises detectors of each of a plurality of groups of radiation
detectors detecting) radiation output from the one or more retarders filtered for
a respective polarisation state by one or more polarisers, the one or more detectors
selectively (and typically separately) detecting polarised radiation conforming to
each of (e.g. the) at least first and second of the said polarised spectral modulation
profiles to thereby provide at least respective first and second polarisation-dependent
radiation intensity measurements from which the presence of the gas can be determined.
[0083] It may be that the apparatus comprises a plurality of groups of radiation detectors,
each of the said groups comprising radiation detectors to detect radiation output
from the one or more retarders filtered for respective polarisation states by one
or more polarisers, the radiation detectors of each of the groups selectively and
separately detecting on different detectors at the same time polarised radiation conforming
to each of at least first and second of the said polarised spectral modulation profiles
to thereby provide at least respective first and second polarisation-dependent radiation
intensity measurements from which the presence of the gas can be determined.
[0084] It may be that the plurality of groups of radiation detectors are configured to selectively
and separately detect radiation filtered for the said respective polarisation states
at the same time as each other.
[0085] It may be that the radiation detectors of the said plurality of groups of radiation
detectors are provided on a common focal plane.
[0086] It may be that the said plurality of groups of radiation detectors are offset from
each other, for example in the common focal plane.
[0087] It may be that the apparatus is configured to spatially modulate the polarisation
of the received radiation by the said plurality of groups of radiation detectors detecting
polarised radiation output from the one or more retarders, the respective groups of
radiation detectors being offset from each other, for example in a common focal plane.
[0088] It may be that the radiation detectors of the said plurality of groups of radiation
detectors are provided in a tessellated arrangement.
[0089] It may be that the plurality of groups of radiation detectors are provided in a periodic
tessellated arrangement, the periodicity of the tessellated arrangement being in respect
of the linear polarisation angles of radiation which the detectors of the groups are
configured to detect. This allows the polarisation of the detected radiation to be
spatially modulated across the tessellated arrangement.
[0090] It may be that the said radiation detectors within each of the said groups are tessellated
in a respective 2x2 or a 2x3 arrangement or a 2x4 arrangement. It may be that the
2x2, 2x3 or 2x4 arrangements are repeating patterns in respect of the linear polarisation
states of the radiation to be detected by the detectors of the respective arrangements.
[0091] It may be that the radiation detectors within each of the said groups are adjacent
to each other.
[0092] It may be that the radiation detectors of a group of the said plurality of groups
are configured to detect radiation output from the one or more retarders filtered
for the same respective linear polarisation states as respective radiation detectors
of one or more or each of the other groups of the plurality of groups.
[0093] It may be that, within each of the groups of radiation detectors, the radiation detectors
are each to detect radiation filtered for a different linear polarisation state from
the other detectors in the respective group.
[0094] It may be that the apparatus comprises, for each of the groups of radiation detectors,
a corresponding group of polarisers for filtering for the respective polarisation
states the radiation output by the one or more retarders, the corresponding group
of polarisers comprising one or more polarisers.
[0095] It may be that each of the said groups of radiation detectors are provided with a
plurality of (typically fixed angle) linear polarisers, each to receive (or it may
be that the method comprises each of the said groups of radiation detectors having
a plurality of (typically fixed angle) linear polarisers each receiving) radiation
having spectrally modulated polarisation from the one or more retarders and output
(or outputting) linearly polarised radiation (e.g. at a respective linear polarisation
angle) to a respective radiation detector of the said plurality of radiation detectors
of the said group, typically to provide respective polarisation dependent intensity
measurements from which presence of the gas can be determined. It may be that the
plurality of linear polarisers for each group comprise linear polarisers of different
linear polarisation angles.
[0096] It may be that the groups of polarisers are provided in a tessellated arrangement.
[0097] It may be that the groups of polarisers are provided in a periodic tessellated arrangement,
the periodicity being in respect of the linear polarisation angles of the polarisers
of the said groups. This helps to enable the spatial modulation of the polarisation
of the detected radiation across the tessellated arrangement.
[0098] It may be that the groups of polarisers are provided on a common focal plane.
[0099] It may be that the groups of polarisers are offset from each other, for example in
the common focal plane.
[0100] It may be that the said groups of radiation detectors are provided adjacent to each
other in a (typically two dimensional) radiation detector array.
[0101] It may be that the apparatus comprises a controller to determine (or it may be that
the method comprises determining), for each of the said groups, the presence of a
gas (preferably the absorption or emission strength of the gas). It may be that the
controller is configured to determine (or it may be that the method comprises determining),
in dependence on the said determinations, an (typically two-dimensional) image selectively
indicating presence of gas (preferably relative absorption or emission strengths of
the gas) at a plurality of locations across the image. It may be that the controller
is configured to determine gas detection measurements based on the respective at least
first and second polarisation-dependent intensity measurements for each of the respective
groups, for example by determining a measure of fractional linear polarisation based
thereon.
[0102] It may be that the controller is configured to reduce or remove systematic errors
(e.g. in the fractional linear polarisation measurement or in the polarisation dependent
radiation intensity measurements from which the presence of gas may be determined)
depending on a combination of polarisation modulations applied to the received radiation
in different domains. It may be that the controller is configured to reduce or remove
systematic errors depending on a plurality of polarisation modulations applied to
the received radiation by the apparatus, the plurality of polarisation modulations
comprising the one or more retarders spectrally modulating the linear polarisation
state of the received radiation to provide the received radiation with a plurality
of polarised spectral modulation profiles which are offset in phase from each other,
each of which is associated with a different linear polarisation angle, and any one
or more of: the one or more retarders spectrally modulating the linear polarisation
state of the received radiation in accordance with different polarised spectral modulation
profiles for different angles of incidence of the radiation on the one or more retarders
(e.g. for radiation of the same linear polarisation angle); the apparatus modulating
the polarisation state of radiation output by the one or more retarders differently
over time; the apparatus spatially modulating the received radiation by the said plurality
of groups of radiation detectors detecting radiation output from the one or more retarders,
each of the groups of detectors comprising radiation detectors selectively and separately
detecting on different detectors at the same time polarised radiation conforming to
each of at least first and second of the said polarised spectral modulation profiles
to thereby provide at least respective first and second polarisation-dependent radiation
intensity measurements from which the presence of the gas can be determined, the respective
groups of radiation detectors being offset from each other, for example in a common
focal plane. It may be that the respective groups of radiation detectors are offset
from each other in a periodic tessellated arrangement, the periodicity of the tessellated
arrangement being in respect of the linear polarisation angles of radiation which
the detectors of the groups are configured to detect.
[0103] Thus, it may be that the one or more retarders are to spectrally modulate the linear
polarisation state of the received radiation to provide the received radiation with
a plurality of polarised spectral modulation profiles which are offset in phase from
each other, each of which is associated with a different linear polarisation angle,
and that the apparatus is to additionally modulate the polarisation of the received
radiation in one or more other domains, such as the angular, temporal and/or spatial
domains. In this case, the polarised spectral modulation profiles to which the radiation
detected by the detectors conform may be provided by the said spectral modulation,
while the additional modulation helps to improve the accuracy with which radiation
conforming to those profiles can be measured. In this case, it may be that the controller
is configured to reduce or remove systematic errors depending on the said additional
modulation.
[0104] As mentioned above, it may be that the groups of radiation detectors, the corresponding
groups of polarisers, or each of the groups of radiation detectors and the corresponding
groups of polarisers are arranged in a respective periodic tessellation arrangement,
the periodicity being in respect of the linear polarisation angles of radiation which
the detectors of the groups are configured to detect or the linear polarisation angles
of the polarisers of the said groups. This periodicity constitutes a polarisation
modulation in the spatial domain. This may be complementary to, for example, the polarisation
modulation provided by the one or more retarders spectrally modulating the linear
polarisation state of the received radiation to provide the received radiation with
a plurality of polarised spectral modulation profiles which are offset in phase from
each other, each of which is associated with a different linear polarisation angle.
In this case, the polarisation modulation provided by the one or more retarders may
provide the received radiation with the polarisation spectral modulation profiles
which allow the presence of gas to be determined, while the additional spatial modulation
helps to improve the accuracy with which radiation conforming to those profiles can
be measured. This polarisation modulation in the spatial domain may provide additional
information which can be used to mitigate systematic errors, for example in the fractional
linear polarisation measurement or in the polarisation dependent intensity measurements
from which the presence of gas may be determined.
[0105] It may be that the controller is to determine (or it may be that the method comprise
determining), for each of a plurality of radiation detectors (e.g. for each radiation
detector of one or more or each of the said groups), whether gas is present in dependence
on polarisation-dependent radiation intensity measurements (i.e. radiation intensity
measurements of polarised radiation) by that radiation detector and by one or more
radiation detectors proximal to that radiation detector. The said one or more radiation
detectors proximal to the said radiation detector may comprise one or more radiation
detectors of a said group comprising the said radiation detector. The said one or
more radiation detectors proximal to the said radiation detector may comprise one
or more radiation detectors of a said group which does not comprise the said radiation
detector.
[0106] It may be that the controller is configured to use a polarisation-dependent radiation
intensity measurement from one of the radiation detectors in the determination of
the presence of the gas for a plurality of (e.g. two or more, three or more or four
or more) positions in the focal plane. It may be that the controller is configured
to use a polarisation-dependent radiation intensity measurement from each of a plurality
of detectors in the determination of the presence of the gas for a respective plurality
of positions in the focal plane. In this way, the spatial resolution of the image
may be increased (for example compared to determining whether the gas is present once
per group of detectors). For example, a determination as to whether gas is present
may be determined in this way for each detector (e.g. for each detector of each group).
In addition, systematic errors which may otherwise appear in the fractional linear
polarisation measurement or in polarisation dependent intensity measurements, for
example due to aliasing and/or due to dead pixels, can be reduced or removed (for
example as compared to determining whether the gas is present once per group of detectors).
For example it may be that the controller is configured to determine whether gas is
present in respect of a first position on a focal plane depending on polarisation-dependent
radiation intensity measurements (i.e. radiation intensity measurements of polarised
radiation) by at least a first one of the radiation detectors and a second one of
the radiation detectors, and to determine whether gas is present in respect of a second
position on the focal plane depending on polarisation-dependent radiation intensity
measurements by at least the first one of the radiation detectors and a third one
of the radiation detectors different from the second one of the radiation detectors.
It may be that the controller is configured to determine whether gas is present in
respect of a first position on a focal plane depending on polarisation-dependent radiation
intensity measurements (i.e. radiation intensity measurements of polarised radiation)
by at least a first one of the radiation detectors, a second one of the radiation
detectors and a third one of the radiation detectors, and to determine whether gas
is present in respect of a second position on the focal plane depending on polarisation-dependent
radiation intensity measurements by at least the first one of the radiation detectors,
the second one of the radiation detectors and a fourth one of the radiation detectors
different from the third one.
[0107] Thus, by spatially modulating the polarisation of the received radiation by the plurality
of groups of radiation detectors detecting radiation output from the one or more retarders,
the respective groups of radiation detectors being offset from each other in a periodic
tessellated arrangement, the periodicity of the tessellated arrangement being in respect
of the linear polarisation angles of radiation which the detectors of the groups are
configured to detect, information from respective proximal (e.g. surrounding) pixels
can be used to determine the fractional linear polarisation, and thus the presence
of the gas, in respect of the location of each of the detectors (e.g. in respect of
the location of each pixel in the array). It may be typically assumed that the gas
concentration varies smoothly, and thus that the fractional linear polarisation varies
smoothly (e.g. at a frequency greater than the individual pixel level, or at a frequency
greater than the spatial modulation frequency) across the radiation detectors of the
plurality of groups (e.g. over the focal plane of the detector array). As such, a
systematic issue with an individual pixel can be mitigated by the spatial modulation.
In the Fourier domain, this can be thought of as considering information at a predefined
spatial modulation carrier frequency (i.e. the frequency with which the polarisation
is spatially modulated across the detectors) and removing or discarding information
at frequencies which differ (or which differ significantly) from the predetermined
spatial modulation frequency (which may be information from individual pixels producing
systematic errors).
[0108] It may be that the polarisation-dependent radiation intensity measurements detected
by the first and second, or first, second and third, or first, second and fourth ones
of the radiation detectors, are determined by detecting polarised radiation conforming
to different ones of the polarised spectral modulation profiles.
[0109] It may be that the apparatus (e.g. the controller) is configured to output the said
two-dimensional image to a user interface such as a display.
[0110] It may be that the said plurality of groups of radiation detectors comprises a first
group arranged to receive (or receiving) radiation incident on the objective lens
at a first angle of incidence and a second group arranged to receive (or receiving)
radiation incident on the objective lens at a second angle of incidence different
from the first angle of incidence. It may be that the first and second angles of incidence
are offset from each other by at least 45°.
[0111] It may be that each of one or more of the said groups, or each of the said groups,
comprise a first detector to detect (or detecting) radiation filtered for a first
linear polarisation angle and conforming to the said first polarised spectral modulation
profile and a second detector to detect (or detecting) radiation filtered for a second
linear polarisation angle and conforming to the said second polarised spectral modulation
profile. It may be that each of one or more of the said groups, or each of the said
groups, comprise a third detector to detect (or detecting) radiation filtered for
a third linear polarisation angle and conforming to a third said polarised spectral
modulation profile. It may be that each of one or more of the said groups, or each
of the said groups, comprise a fourth detector to detect (or detecting) radiation
filtered for a fourth linear polarisation angle and conforming to a fourth said polarised
spectral modulation profile.
[0112] It may be that the apparatus comprises a controller to derive (or it may be that
the method comprises deriving) a concentration of the gas in the atmosphere from the
polarisation-dependent intensity measurements by the first and second groups of radiation
detectors.
[0113] It may be that the received radiation comprises or consists of radiation emitted
by the sun (and typically scattered by the Earth's atmosphere).
[0114] It may be that the received radiation received by the one or more retarders has passed
through a gas measurement volume within which the gas is to be detected.
[0115] It will be understood that, within the scope of the appended claims, optional and
preferred features of each aspect of the present disclosure are also optional or preferred
features of each other aspect of the present disclosure where appropriate. For example,
it may be that apparatus features correspond to method features.
BRIEF DESCRIPTION OF THE DRAWINGS
[0116] Embodiments of the disclosure are further described hereinafter with reference to
the accompanying drawings, in which:
Figure 1a schematically illustrates an apparatus to detect polarised radiation from
which the presence of a gas can be determined, wherein the apparatus includes a telecentric
lens;
Figure 1b schematically illustrates an apparatus to detect polarised radiation from
which the presence of a gas can be determined, wherein the apparatus includes a telecentric
lens and a birefringent filter;
Figure 2 schematically illustrates how the absorption spectrum of a gas compares to
intensity signals corresponding to the polarised spectral modulation profiles for
linear polarisation angles of 0° and 90°;
Figure 3 schematically illustrates an example calibration curve to determine the absorption
strength of the gas from the fractional linear polarisation of detected polarised
radiation;
Figure 4a schematically illustrates an apparatus to detect polarised radiation from
which the presence of a gas can be determined, wherein the apparatus includes a telecentric
lens and a rotating polarisation filter;
Figure 4b is a similar diagram to Figure 4a, but with a rotating half waveplate and
fixed linear polarisation angle polariser in place of the rotating polarisation filter;
Figure 5a schematically illustrates a tiled array of radiation detectors to detect
four different polarisations of radiation;
Figure 5b schematically illustrates an apparatus to detect polarised radiation from
which the presence of a gas can be determined, wherein the apparatus includes groups
of four radiation detectors with corresponding groups of four micropolarisers provided
over the radiation detectors of a corresponding group of radiation detectors;
Figure 6a is a plot showing an example of how a portion of the absorption spectrum
of NO2 matches the transmission response of a pre-filter comprising a bandpass interference
filter and a one stage birefringent filter and the intensity signals corresponding
to polarised spectral modulation profiles for linear polarisations of 0°, 45°, 90°
and 135° for temperatures of the retarders of 10°C, 20°C and 30°C, the retarders being
provided in an athermal configuration;
Figure 6b illustrates a similar plot to Figure 6a but where the pre-filter comprises
a bandpass interference filter and a two-stage birefringent filter and the birefringent
elements of the retarders, instead of being provided in an athermal configuration,
are made from the same material (quartz);
Figure 6c shows plots of determined fractional linear polarisation with respect to
a zero target gas concentration reference at a fixed temperature (e.g. 20°C) versus
NO2 gas concentration for temperatures of 10°C, 20°C and 30°C wherein the birefringent
elements of the apparatus are arranged in an athermal combination;
Figure 6d shows plots of determined fractional linear polarisation with respect to
a temperature dependent zero target gas concentration reference for apparatus temperatures
of 10°C, 20°C and 30°C, and of determined fractional linear polarisation with respect
to a 20°C zero target gas concentration reference versus NO2 gas concentration for apparatus temperatures of 10°C, 20°C and 30°C;
Figure 7 schematically illustrates an optical set-up to detect polarised radiation
from which the presence of a gas can be determined using a non-telecentric lens; and
Figure 8 is a flow chart schematically illustrating a method for determining the presence
of a gas.
DETAILED DESCRIPTION
[0117] An apparatus and method for determining the presence of a gas are provided. The apparatus
and method may be for determining the presence of a target gas within a volume of
unknown gas composition. The apparatus and method may further be for determining the
radiation absorption or emission strength of a gas, for example a radiation absorption
or emission strength relative to a calibration reference or threshold. The apparatus
and method may be for determining a concentration of the gas. The gas may be any gas
with a known absorption or emission spectrum. The gas may be an atmospheric trace
gas. Example gases that may be determined to be present may include NO
2, O
3, CO, CO
2 and CH
4, however the apparatus is not limited to detect only these gases and these are mentioned
for example only. The apparatus to determine the presence of gas may be used for example
in the remote sensing of atmospheric trace gases for air pollution monitoring and
source detection. The apparatus may be used in ground-based air pollution monitoring
or targeted pollution measurements from a mobile platform, for example in a law-enforcement
campaign and using measurements from drones. The apparatus may also, or alternatively,
be used in gas leak detection, for example in the oil and gas industry using a hand-held
system, drones or satellites. The apparatus may also or alternatively be used in earth
observation from space, for example NO
2 measurements may be performed using cubesats, or CO
2 localisation may be measured.
[0118] An apparatus and method are provided for determining the presence of a gas which
may be compact and robust, making it suitable for use in environments as described
above.
[0119] Although the following description assumes that the presence of the gas is determined
by way of absorption of radiation by the gas, it will be understood that the techniques
and apparatus described herein are applicable to determining presence of the gas by
emission of radiation by the gas.
[0120] A first embodiment will be described which relates to Figures 1a, 1b and 2. Figure
1a schematically illustrates apparatus 100 for detecting presence of a gas. The apparatus
100 may comprise a telecentric objective lens 102 for capturing and converging incoming
electromagnetic radiation 101 (which may be from a broadband radiation source) which
has passed through a gas detection volume 103 which may comprise the gas to be detected
(e.g. such that selected frequencies of the radiation may have been attenuated due
to absorption by the gas in accordance with its gas absorption spectrum). As the lens
102 is telecentric, the optical axis direction of the converging beam is independent
of position inside the field of view of the lens 102. The received radiation 101 may
comprise sunlight (which is typically scattered by the atmosphere) and the gas detection
volume 103 may comprise a column volume of the earth's atmosphere through which the
sunlight propagates. The gas to be detected may be a trace gas in the earth's atmosphere.
Alternatively, the received radiation may comprise radiation from a local electromagnetic
radiation source (e.g. an incandescent light source or light emitting diode (LED))
and the gas detection volume may comprise or consist of a local (e.g. closed) gas
detection volume. In the former case, the received radiation is typically provided
with a polarisation pattern as a result of scattering by air, water and aerosol molecules
in the atmosphere. In the latter case, the radiation may be unpolarised or it may
be linearly polarised, for example at a linear polarisation angle of 0°. The received
radiation typically comprises electromagnetic radiation of a plurality of electromagnetic
frequencies across a frequency range. Typically the frequencies of received radiation
extend over a continuous range of frequencies. Typically the radiation may comprise
any one or more of ultraviolet, visible, near-infrared, mid-infrared electromagnetic
radiation.
[0121] Converging radiation 105 output by the telecentric objective lens 102 may be directed
towards an optical filter 104. The optical filter 104 may be a bandpass interference
filter for selectively transmitting electromagnetic radiation within a selected frequency
range defined by a passband. The passband may be selected to allow electromagnetic
radiation of frequencies corresponding to a portion of the absorption spectrum of
the gas to be detected. The portion of the absorption spectrum may comprise periodic
or quasi-periodic absorption lines, that is, corresponding to a portion of the absorption
spectrum of the gas to be detected comprising a plurality of absorption lines having
respective absorption peaks which are offset from each other in frequency, the frequency
spacing between successive ones of the said peaks being substantially equal.
[0122] Figure 2 illustrates a portion 107 of an exemplary absorption spectrum which comprises
four substantially periodic absorption peaks 201, adjacent ones of which are offset
from each other in frequency by a substantially constant frequency spacing Δf
1 210. The portion 107 of the absorption spectrum is selected by the optical filter
104 having a passband 200 which transmits radiation having frequencies corresponding
to the portion 107 of the absorption spectrum.
[0123] Referring back to Figure 1a, the radiation transmitted by the filter 104 may be output
to a spectral modulator 115. In the exemplary embodiment, the spectral modulator 115
comprises a linear polariser 106 and one or more retarders 108, the one or more retarders
108 in the example of Figure 1 comprising a retarder plate 109 and a quarter waveplate
110. It will be understood that if the received radiation 101 is linearly polarised,
for example at a known angle of linear polarisation, the linear polariser 106 may
not be required. In some embodiments, the quarter waveplate 110 may be omitted; this
will be explained in more detail below. The linear polariser 106, where provided,
may linearly polarise radiation transmitted by the filter 104, for example at a linear
polarisation angle of 0°.
[0124] The one or more retarders 108 receive linearly polarised, converging radiation at
a linear polarisation angle of 0° from the polariser 106 (or from the filter 104 if
the polariser 106 is omitted). The spread of angles due to the converging beam is
identical for all points in the field of view of the lens 102. Accordingly the angle
of incidence of the radiation on the one or more retarders is independent of its angle
of incidence on the telecentric objective lens 102. The one or more retarders 108
may comprise one or more non-birefringent retarders such as a Fresnel Rhomb, but more
typically the retarders 108 may comprise or consist of one or more birefringent elements.
In the following description it will be assumed that the one or more retarders 108
comprise one or more birefringent elements.
[0125] The birefringent elements (typically birefringent crystals) of the one or more retarders
109 each have slow and fast axes having different refractive indices for polarised
components of received radiation parallel and perpendicular to the optic axis 111.The
optic axis 111 may be oriented at 45° relative to the linear polarisation angle (0°)
of the radiation it receives from the polariser 106 or filter 104. When the radiation
enters the one or more birefringent elements, a first portion of the radiation propagates
along the fast axis and a second portion propagates along the slow axis before the
two portions recombine at the output side of the birefringent elements. The difference
in the refractive indices of the slow and fast axes causes an optical path length
difference between the optical paths travelled by the first and second portions of
the radiation. This retardance modulates the polarisation state of the radiation.
The way in which the polarisation state of the radiation is modulated depends on the
frequency of the radiation. Because the radiation comprises a range of different frequencies,
the radiation output by the one or more birefringent elements therefore has a number
of components of different polarisation.
[0126] The polarisation state of the recombined radiation output by the birefringent elements,
and thus the amplitudes of respective components of the polarisation state at particular
linear polarisation angles, depend on the phase difference between the component of
the radiation propagating on the fast axis and the component of the radiation propagating
on the slow axis. This phase difference is frequency dependent. Accordingly, for some
frequencies, a component of the recombined radiation at a particular linear polarisation
angle has a maximum amplitude, while for other frequencies the component of the recombined
radiation at that linear polarisation angle has a minimum amplitude, and for other
frequencies the component of the recombined radiation at that linear polarisation
angle will have an amplitude between the minimum and maximum amplitudes. In this way,
as a function of frequency, the amplitude of the component of the radiation at that
linear polarisation angle may be spectrally modulated (typically substantially sinusoidally)
in accordance with a polarised spectral modulation profile comprising a plurality
of substantially periodic maxima and minima. Indeed, the polarisation state of the
radiation may be spectrally modulated in accordance with different polarised spectral
modulation profiles for different angles of linear polarisation. The spectral modulation
profiles may be offset from each other in phase corresponding to the relative difference
in the linear polarisation angle at which they are filtered (the phase difference
being twice the difference in linear polarisation angle because all linear polarisation
phenomena have a 180° ambiguity and thus periodicity).
[0127] Typically the thickness 150 of the one or more birefringent elements is selected
to provide a retardance which converts the polarisation state of the received radiation
at different frequencies to a polarisation state having components of linear polarisation
(and thus polarised spectral modulation profiles) at linear polarisation angles of
0° and 90° and components which are right and left hand circularly polarised, but
with no components of linear polarisation at +/-45°. The quarter waveplate 110 (where
provided), which may have an optic axis arranged parallel to the linear polarisation
state of the radiation received by the one or more birefringent elements, converts
the polarisation state of the radiation from the retarder 109 into radiation which
has components of linear polarisation (and thus polarised spectral modulation profiles)
at linear polarisation angles of 0°, 45°, 90° and 135°, thereby providing radiation
having a full state of linear polarisation.
[0128] In a first example, the quarter waveplate 110 is omitted and radiation output by
the retarder plate 109 having spectrally modulated polarisation is detected by first
and second detectors 142, 144 of detection apparatus 112, the detectors being provided
in this example on a common focal plane 146. A first linear polariser 147 which linearly
polarises radiation at a linear polarisation angle of 0° (although this 0° angle is
arbitrary, it may act as a 0° reference point for the rest of the optics of the apparatus)
is provided over the first detector 142 and a second linear polariser 148 which linearly
polarises radiation at a linear polarisation angle of 90° is provided over the second
detector 144. It may be that the first and second linear polarisers 147, 148 each
receives from the one or more retarders radiation having a plurality of linear polarisation
components and provides at a single output radiation having a single predefined linear
polarisation, in this case for example 0° and 90° for the first and second polarisers
147, 148 respectively. The first detector 142 may detect the polarised radiation output
from the first linear polariser 147 (indiscriminately with frequency over the selected
frequency range) to provide a first polarisation dependent intensity measurement I
0. The second detector 144 may detect the polarised radiation output from the second
linear polariser 148 (indiscriminately with frequency over the selected frequency
range) to provide a second polarisation dependent intensity measurement I
90.
[0129] The radiation output from the one or more retarders 108 comprises radiation having
polarisation spectrally modulated in accordance with the said plurality of polarised
spectral modulation profiles in a common beam of radiation. The radiation filtered
by polariser 147 for linearly polarised radiation at 0° is filtered from a beam of
radiation comprising radiation spectrally modulated in accordance with both the first
and second said polarised spectral modulation profiles (and the radiation filtered
by polariser 147 for linearly polarised radiation at 90° is filtered from a beam of
radiation comprising radiation having polarisation spectrally modulated in accordance
with the first and second said polarised spectral modulation profiles). By providing
radiation having polarisation spectrally modulated in accordance with the polarised
spectral modulation profiles in a common beam of radiation, the radiation modulated
in accordance with each of the polarised spectral modulation profiles follows substantially
the same optical path from the one or more retarders 108 to the detectors 142, 144.
Thus, by splitting the beam in the polarisation domain by way of the retarders, rather
than splitting the beam into two portions which each follow a separate physical optical
path (as would be the case in a classical interferometer such as a Michelson interferometer
which splits the beam into two independent arms before recombining them or a Mach-Zender
interferometer which also utilises more than one optical path), the different polarisation
states of the beam follow a common optical path. As such, the apparatus is only sensitive
to relative temperature effects on the different polarisation states and not on temperature
dependent path length variations (as would be the case in a classical interferometer).
This helps to keep the apparatus compact and makes the apparatus more tolerant to
temperature changes. The relative temperature effects on the different polarisation
states can be accounted for in different ways, as will be described in more detail
below. Due to the common optical path followed by the different polarisation states
of the beam, the apparatus is also more tolerant to vibration.
[0130] When the radiation having the spectrally modulated polarisation is filtered at a
polarisation angle corresponding to a respective said polarised spectral modulation
profile, the intensity of the detected radiation varies with frequency of radiation
in accordance with that profile. This is illustrated in Figure 2 which shows first
and second signals 204, 206. Although signals 204, 206 are shown offset from each
other in amplitude in Figure 2, it will be understood that this is only for illustration
purposes and that both signals in fact vary between normalised amplitudes of 0 and
1. The first signal 204 is the variation of intensity with frequency of radiation
output by the spectral modulator 115 and filtered for a linear polarisation angle
of 0° by the first linear polariser 147 provided over the first detector 142.
[0131] The second signal 206 is the variation of intensity with frequency of radiation output
by the spectral modulator 115 and filtered for a linear polarisation angle of 90°
by the second linear polariser 148 provided over the second detector 144. As there
is a 90° difference between the linear polarisation angles of the first and second
polarisers 147, 148, the first and second signals 204, 206 are offset from each other
in phase by 180°.
[0132] By the polarised spectral modulation profiles being provided in a common beam of
radiation, and the radiation detectors 142, 144 selectively, separately and simultaneously
detecting polarised radiation conforming to the respective polarised spectral modulation
profiles, the apparatus can be made with few (if any) moving parts, enabling a robust
construction. In addition, more sensitive and accurate gas detection can be achieved
at least because the respective polarisation dependent radiation intensity measurements
are made under identical temperature, vibration, optical alignment and gas concentration
conditions.
[0133] The thicknesses of the one or more retarders 108 may be selected such that the (substantial)
periodicity of the transmission maxima and minima of the polarised spectral modulation
profiles substantially matches the period or quasi-period of the absorption peaks
of the said portion 107 of the absorption spectrum of the gas. In addition, the retarders
108 and polariser 106 (where provided) may be rotated until the maxima 250 of the
first polarised spectral modulation profile (and thus of the first signal 204) are
aligned (and substantially overlap) in frequency with the absorption lines, and preferably
the absorption peaks 201, of the portion 107 of the absorption spectrum of the gas
in the selected frequency range, and the transmission minima 252 of the first polarised
spectral modulation profile (and thus of the first signal 204) are aligned (and substantially
overlap) in frequency with portions 203 of the absorption spectrum of the gas between
the absorption peaks 201 (preferably off the absorption lines). By virtue of the 180°
phase difference between the first and second spectral modulation profiles, the transmission
minima 260 of the second polarised spectral modulation profile (and thus of the second
signal 206) are thus aligned (and overlap) in frequency with the absorption peaks
201 of the portion 107 of the absorption spectrum of the gas, and the transmission
maxima 262 of the second polarised spectral modulation profile (and thus of the second
signal 206) are aligned (and overlap) in frequency with portions 203 of the absorption
spectrum of the gas between the absorption peaks 201 in the selected frequency range.
In Figure 2, any effect of absorption of radiation by the gas is omitted from the
first and second signals 204, 206 for illustration purposes. However, it will be understood
that when the gas is present a portion of the resulting spectrum has a lower intensity
at the spectral modulation maxima due to absorption of radiation by the gas.
[0134] However, because the transmission maxima 262 of the second signal 206 are out of
phase with the absorption peaks 201 of the gas, the second polarisation dependent
intensity measurement I
90 will be relatively unaffected by the presence of the gas. Accordingly, by comparing
the first and second polarisation dependent intensity measurements, the presence of
the gas can be determined. This may be done by a controller 120 which receives radiation
detection data from the first and second detectors 142, 144 relating to the first
and second polarisation dependent intensity measurements, I
0, I
90. The controller 120 may comprise data acquisition circuitry for receiving the radiation
detection data and processing circuitry for processing the received radiation detection
data (e.g. by executing stored computer program instructions causing it to process
the received radiation detection data) to thereby compare the first and second polarisation
dependent intensity measurements I
0, I
90 to thereby detect presence of the gas.
[0135] The controller 120 may determine the presence of the gas (and in some cases the concentration
of the gas) by determining fractional linear polarisation data relating to the fractional
linear polarisation of the radiation output from the one or more retarders 108 based
on the polarisation dependent intensity measurements. More specifically in this case,
the controller 120 may determine the presence of the gas from the first and second
intensity measurements I
0, I
90 by determining fractional linear polarisation data relating to the fractional linear
polarisation of the light output from the one or more retarders 108 from the first
and second polarisation dependent intensity measurements obtained by the first and
second detectors 142, 144 I
0, I
90. This will be explained as follows.
[0136] A Stokes vector [
Imeas, Q
meas,
Umeas, Vmeas]
T may be used to describe the linear polarisation state of the radiation after it has
been spectrally modulated by the spectral modulator 115, and this may be defined as
below:

where: I
45 would be a third polarisation dependent intensity measurement by a radiation detector
having a linear polariser provided in front of it, the linear polariser transmitting
radiation of linear polarisation angle of 45°;
I135 would be a fourth polarisation dependent intensity measurement by a radiation detector
having a linear polariser provided in front of it, the linear polariser transmitting
radiation of linear polarisation angle of 135°;
IRHC would be a fifth polarisation dependent intensity measurement by a radiation detector
having a polariser provided in front of it, the polariser transmitting radiation of
right hand circular polarisation; and
ILHC would be a sixth polarisation dependent intensity measurement by a radiation detector
having a polariser provided in front of it, the polariser transmitting radiation of
left hand circular polarisation.
[0137] In this example it is assumed that the U component of the Stokes vector is zero because
the retarder plate 109 is designed not to output radiation polarised at 45° or 135'.
Accordingly, radiation is only measured, at the detection apparatus 112, at linear
polarisation angles of 0° and 90°. This is indicated by U
meas and V
meas being set equal to zero in equation 2 below, but it will be understood that U
meas and V
meas are not typically physically measured in this embodiment.

[0138] Using Equation 2, the controller 120 may determine the values of
Imeas and
Qmeas based on the first intensity value, I
0, and the second intensity value, I
90. Assuming that the U and V components of the Stokes vector are zero, the fractional
linear polarisation can be determined from:

and the Angle of Linear Polarisation, AoLP, can be assumed to be 0. In this case,
it may be that the fractional linear polarisation data comprises a measure of the
fractional linear polarisation of the radiation output by the spectral modulator 115.
[0139] The fractional linear polarisation of the emergent radiation from the spectral modulator
115 is directly proportional to the absorption strength of the gas. A calibration
reference may be used to convert the fractional linear polarisation into absorption
strength. An example calibration graph to determine absorption strength is illustrated
in Figure 3. The absorption strength of the gas may be found by determining the fractional
linear polarisation of the radiation detected at the detection apparatus 112 and matching
it to the absorption strength of the gas that lies on the calibration curve at that
value of fractional linear polarisation.
[0140] The steeper the calibration curve 302 of Figure 3, the more sensitive the determined
fractional linear polarisation 304 is to absorption strength 306. This is determined
at least partially by how close the frequency spacing 210 between the absorption peaks
201 of the portion of the absorption spectrum of the gas 107 matches the frequency
spacing 212 between maxima 250 of the first signal 204. In the present embodiment,
the sensitivity also depends on the alignment (or overlap) between the maxima of the
first signal 204 and the absorption peaks of the gas, and on the alignment (or overlap)
between the maxima of the second signal 206 and portions of the absorption spectrum
off the said peaks, in the selected frequency range. During assembly of the apparatus,
the angles and thicknesses of the birefringent elements in the spectral modulator
115 may be adjusted until this line 302 is as steep as possible.
[0141] A calibration graph 300, as illustrated in Figure 3, may be determined by using an
enclosed control gas measurement volume. A known concentration of the gas having a
known absorption strength 306 may be provided in an enclosed volume and radiation
may be transmitted through the gas from a known radiation source, such as an incandescent
bulb. The radiation may travel through a known path length of the enclosed volume
and be detected using the apparatus 100. The fractional linear polarisation 304 may
be determined from first and second polarisation dependent intensity measurements
as described above, by the controller 120 using Equation 2. This may be repeated for
multiple known concentrations (and thus absorption strengths 306) of the gas. A calibration
graph 300 may be formed relating the fractional linear polarisation 304 to the known
absorption strengths 306 of the gas. A concentration of the gas may be determined
from the determined absorption strength of the gas. Additionally, or alternatively,
the concentration of the gas may be determined from the fractional linear polarisation
using a calibration graph of the concentration of the gas against the fractional linear
polarisation, which may be determined in substantially the same way.
[0142] While in theory it may be expected that the sensitivity of the apparatus would be
improved if a single absorption feature of the absorption spectrum of the gas (rather
than a plurality of absorption features as shown) was to be interrogated, in practice
the achievable signal to noise ratio has been found to be greater when multiple absorption
features are interrogated with the same measurement.
[0143] The alignment (or overlap) in frequency of the transmission maxima and minima of
the first and second signals 204, 206 with the absorption peaks 201 and the portions
of the absorption spectrum of the gas between absorption peaks (as the case may be)
within the selected frequency range may be affected by the temperature of the one
or more birefringent elements. This may be due to the temperature dependence of the
birefringence of the birefringent elements and due to the thermal expansion of the
birefringent material which changes the path length through the material. This temperature
dependence can adversely affect the determination of the presence of the gas, and
cause errors in the gas concentration measurements.
[0144] This temperature sensitivity can be inhibited by the one or more birefringent elements
comprising a plurality of birefringent elements of different types which are provided
in a subtractive combination which cancels out temperature dependency of the birefringent
elements. For example, it may be that the plurality of birefringent elements comprises
first and second birefringent elements which have different birefringences. It may
be that the fast axis of the first said birefringent element is aligned with the slow
axis of the second birefringent element. It may be that the change with temperature
of the retardance of the first said birefringent element is substantially equal to
the change with temperature of the retardance of the second said birefringent element
such that they cancel out. As the first and second elements have different birefringences,
the retardances of the two elements do not cancel out. This helps to maintain alignment
(or overlap) in frequency of the transmission maxima and minima of the first and second
signals 204, 206 with the absorption peaks 201 and the portions of the absorption
spectrum of the gas between absorption peaks (as the case may be) within the selected
frequency range, thereby helping to maintain the accuracy of the apparatus in determining
presence and absorption strength of the gas without having to maintain the apparatus
at a constant temperature. This is discussed further below with reference to Figure
6a.
[0145] Although the phase difference between the first and second signals 204, 206 (and
thus the phase difference between the first and second polarised spectral modulation
profiles) is described above as being 180°, it will be understood that this precise
phase difference (although preferable) is not essential. Different phase differences
will in most cases permit the gas to be detected, albeit at a reduced sensitivity.
The closer the phase difference is to 180°, the more sensitive the gas detection.
Preferably, the phase difference is sufficiently close to 180° that the fractional
linear polarisation is at least 5% of the average (e.g. mean) relative absorption
depth of the absorption lines of the gas within the selected frequency range.
[0146] The temperature sensitivity of the birefringent elements may in some cases alternatively
be compensated for by performing calibrations at a range of temperatures, measuring
the temperature when intensity measurements are made, and applying the appropriate
temperature dependent calibration. However, in this case, typically at least three
polarisation dependent intensity measurements are required. This is explained in more
detail below, particularly with reference to Figures 6c, 6d.
[0147] In a second example, the quarter waveplate 110 is provided such that the radiation
output from the spectral modulator has polarised components at linear polarisation
angles of 0°, 45°, 90°, 135° (i.e. a full state of linear polarisation). In this case,
it can no longer be assumed that the U component of the Stokes vector is zero, and
the fractional linear polarisation may be determined by: determining values for Q
meas, U
meas and I
meas; determining Q
meas/I
meas and U
meas/I
meas from the determined values of Q
meas, U
meas and I
meas; and rotating the [Q
meas, U
meas] co-ordinate system to provide a rotated co-ordinate system [Q
meas', U
meas'] whereby Q
meas'/I
meas>0 and U
meas'/I
meas=0. In this case, Q
meas'/I
meas is indicative of the fractional linear polarisation of the radiation output by the
spectral modulator 115. The Angle of Linear Polarisation (AoLP) may be determined
from the measured Stokes Parameters of the measured Stokes Vector using Equation 3:

[0148] Thus, in such embodiments, the controller 120 may determine the fractional linear
polarisation (and optionally the AoLP) using determined values of
Imeas, Umeas and
Qmeas. In order to determine the Stokes parmeters, rather than two polarisation dependent
intensity measurements being made which correspond to radiation output from the spectral
modulator 115 filtered for linear polarisation angles of 0° and 90°, at least three
polarisation dependent intensity measurements are made (in view of there being three
unknowns, Q, U and I). In this example, three polarisation dependent intensity measurements
are made, typically by respective radiation detectors, typically simultaneously, which
correspond to radiation output from the spectral modulator 115 filtered for linear
polarisation angles of 0°, 60° and 120° by respective linear polarisers which respectively
filter the radiation output by the spectral modulator at linear polarisation angles
of 0°, 60° and 120°. The intensity of radiation filtered for linear polarisation angles
of 0°, 60° and 120° varies in accordance with first, second and third polarised spectral
modulation profiles of the radiation output by the spectral modulator which are offset
from each other in phase by 120° (thereby providing relative phases of 0°, 120° and
240°).
[0149] The first, second and third intensity signal measurements can be described using
Stokes parameters. As there may be no circular polarisation it is assumed that the
V component of the Stokes vector is zero and the Stokes vector [
Imeas, Qmeas,
Umeas]
T may be used to describe the complete polarisation of the radiation output from the
spectral modulator 115 as described below in Equations 4 and 5:

[0150] I
0, I
60 and I
120 may be the first, second and third polarisation-dependent intensity signal measurements
filtered at linear polarisation angles of 0°, 60° and 120° respectively by respective
linear polarisers.
[0151] From Equation 5,
Imeas, Qmeas,
Umeas may be determined and the fractional linear polarisation (and optionally the angle)
of linear polarisation may be calculated as described above. The fractional linear
polarisation may then be converted into an absorption strength or concentration of
the gas by using a pre-determined calibration curve as previously described. Thus,
again in this case, the fractional linear polarisation data may comprise a measure
of the fractional linear polarisation of the radiation output by the spectral modulator
115.
[0152] Because the quarter wave plate 110 provides the radiation output by the spectral
modulator 115 with a full state of linear polarisation, the modulation phase depends
linearly on the angle of the filtering linear polarisers. In addition, as described
above, the three polarisation-dependent intensity measurements allow the full fractional
linear polarisation of the radiation output by the spectral modulator 115 to be determined.
Accordingly, it is not necessary in this example for the maxima of any of the first,
second and third polarised spectral modulation profiles to be aligned (or overlap)
with the absorption lines of the gas in the selected frequency range or for the maxima
of any of the first, second and third polarised spectral modulation profiles to be
aligned (or overlap) off the absorption lines of the gas in the selected frequency
range. This helps to make the apparatus 100 even less sensitive to changes in temperature
as spectral shifts in the polarised spectral modulation profiles can be better tolerated.
Put another way, for the general case when the period of the polarised spectral modulation
profiles is known but not their modulation phases (e.g. due to temperature effects),
at least three polarisation dependent intensity measurement signals are needed to
determine the fractional linear polarisation.
[0153] Although the first, second and third polarised spectral modulation profiles may be
offset in phase from each other by substantially 120° in this embodiment to help provide
maximum sensitivity, the profiles may be offset from each other by any phase. The
more distributed the relative phases of the first, second and third of the said polarised
spectral modulation profiles throughout a 360° cycle, the more efficiently the fractional
linear polarisation can be determined from the first, second and third polarisation-dependent
radiation intensity measurements and hence an offset of 120° may be most efficient.
[0154] In a further embodiment, radiation output from the spectral modulator 115 filtered
for respective linear polarisation states corresponding to four polarised spectral
modulation profiles may be detected, typically by respective detectors, typically
simultaneously, the four polarised spectral modulation profiles being offset in phase
from each other (similar to a four phase detection scheme in interferometry). Figure
6a illustrates an example portion of the absorption spectrum 602 of NO
2 gas for different concentrations of NO
2 gas, and four signals 604-607 showing how the intensities of radiation output by
the spectral modulator 115 and filtered for four respective linear polarisation angles
at 0°, 45°, 90° and 135° varies in accordance with four respective polarised spectral
modulation profiles. The first, second, third and fourth signals 604-607 (and corresponding
polarisation-dependent spectral modulation profiles) which are adjacent to each other
in phase are offset in phase from each other in phase by substantially 90°, the four
signals 604-607 having relative phases of 0°, 90°, 180° and 270°. The frequency spacing
between successive maxima of the polarisation-dependent spectral modulation profiles
may be substantially equal to the frequency spacing between a pair of absorption peaks
of the gas.
[0155] The Stokes parameters can be determined from the four polarisation dependent intensity
measurements as follows. As there may be no circularly polarised radiation output
by the spectral modulator 115, it is assumed that the V component of the Stokes vector
is zero and the Stokes vector [
Imeas, Qmeas,
Umeas]
T may be used to describe the complete polarisation of the radiation output from the
spectral modulator 115 as described below in Equations 6 and 7:

[0156] Where I
0, I
90, I
45 and I
135 may be the first, second, third and fourth polarisation-dependent intensity signal
measurement filtered at linear polarisation angles of 0°, 45°, 90° and 135° respectively.
The fractional linear polarisation, and optionally the angle of linear polarisation,
may then be determined by the controller 120 as previously described and an absorption
strength or concentration of the gas may be determined as described previously. Thus,
again, the fractional linear polarisation data may comprise a measure of the fractional
linear polarisation of the radiation output by the spectral modulator.
[0157] As before, the fractional linear polarisation of the emergent radiation from the
spectral modulator 115 may be directly proportional to the strength of spectral features
of the incident radiation. As radiation is detected at four different linear polarisation
angles, the complete state of linear polarisation of the radiation may be determined
and as a result there is no requirement for any of the four polarised spectral modulation
profiles to be in phase or out of phase with the absorption features.
[0158] Detecting a first, second, third and fourth polarisation-dependent intensity measurement
filtered at linear polarisation angles of 0°, 45°, 90° and 135° is the most efficient
configuration to measure linear polarisation with four measurements (assuming that
the transmissions are equal for all four angles; it will be understood that the 0°
reference is still arbitrary). Many other combinations of linear polarisation angles
may still allow the fractional linear polarisation (and optionally angle of linear
polarisation) to be determined, albeit with reduced efficiency and sensitivity. Preferably,
the fractional linear polarisation determined by the controller 120 is at least 5%
of the average (e.g. mean) relative absorption depth of the absorption lines of the
gas within the selected frequency range.
[0159] Although the examples in which at least three polarisation-dependent intensity measurements
are obtained do not require alignment in frequency between maxima of polarised spectral
modulation profiles to which they conform and absorption lines of the absorption spectrum
of the gas (or with off-peak portions of the absorption spectrum), it can still be
advantageous to implement an athermal retarder arrangement (i.e. comprising a combination
of multiple retarder elements of different materials wherein the temperature dependences
of their retardances cancel out) in order to maximise the accuracy of the gas concentration
measurements which are possible with the apparatus, for example by reducing or eliminating
measurement errors caused by temperature fluctuations causing corresponding fluctuations
in retardances of the one or more retarders.
[0160] As illustrated in Figure 1b, the optical filter 104 of the apparatus of Figure 1a
may comprise a bandpass interference filter 121 and one or more birefringent filter
stages 125. Each of the birefringent filter stages 125 may comprise a rotatable polariser
122, a retarder 126 and optionally a rotatable quarter waveplate 124. The retarder
126 may be a birefringent plate. The dashed lines on the optical elements demonstrate
example optic axes of the optical elements, but it will be understood that these are
examples only and different orientations of the optical axes of the optical elements
are possible. The curved arrows indicate that the optical elements (including the
polarisers, wave plates and retarders), or the optic axes of the optical elements,
may be rotated, typically independently of each other, which is unlike a traditional
Lyot filter. During assembly/tuning of the apparatus, the optic axes of the quarter
wave plate 124, retarder 126 and polariser 122 may initially be aligned with the optic
axis of the polariser 106 of the spectral modulator 115. However, the optic axis of
the polariser 122 can be rotated (e.g. by rotating the polariser) to allow the transmission
response of the birefringent filter 125 to be tuned to the profile of the bandpass
interference filter 121. The optic axis of the polariser 106 can be rotated to tune
the phases of the spectral modulation profiles to which the radiation detected by
the detectors conform to the polariser angles of the detector, given a retardance
of birefringent plate 108. The optical filter 104 is not limited to this configuration
and may comprise any combination of rotatable and non-rotatable polarisers, wave plates
and retarders. The arrows to indicate radiation travelling through the optical components
have not been included in Figure 1b for clarity, however the radiation may follow
a similar path to that shown by the arrows in Figure 1a. The rotational orientation
and the thickness of the retarders and optical elements in the filter 104 may be chosen
to refine the passband 200 of the bandpass filter 121 to provide a combined response
127 having deep minima of transmission at the edges of the passband of the bandpass
filter 121, as illustrated in Figures 2 and 6a.
[0161] Converging radiation from the telecentric objective lens 102 may be incident on the
bandpass filter 121 which transmits radiation with frequencies within the passband
200 of the bandpass filter 121. The filtered radiation from the bandpass filter 121
is then incident on the polariser 122 (which, after rotation to tune the response
of the birefringent filter to the bandpass filter, is typically fixed). The polariser
122 transmits radiation of a certain linear polarisation angle to the quarter waveplate
124 and birefringent plate 126. As described in relation to the spectral modulator
115 above, the position of the modulation depends on the polarisation angle because
of the quarter wave plate 124, thereby allowing a tuneable filter to be implemented.
Although the order of the components in the birefringent filter 125 is reversed in
comparison to the spectral modulator 115, this is merely for ease of alignment. A
single birefringent stage 125 will provide a sinusoidal response which is typically
selected to have a larger period than the response of the spectral modulator 115.
This is illustrated in Figure 6a. However, if a plurality of birefringent stages 125
are provided, a top-hat response can be obtained. This is illustrated in Figure 6b
which shows an example response 127 when two birefringent filter stages are provided.
Accordingly, the birefringent stage(s) have the effect of refining the passband 200
of the bandpass interference filter 121 to provide the combined response 127. The
polarisation angle of the polariser 122 may be different from that of the polariser
106 to provide the filter 104 with the desired response.
[0162] As a result of temperature dependency of the birefringent elements in the retarders
and the birefringent filter 125, the response of the birefringent filter 125 and the
spectral modulation phases (and thus the AoLP) of the polarised spectral modulation
profiles may depend on temperature.
[0163] As discussed above, the apparatus may be made athermal by using an athermal crystal
configuration in which, for example, the birefringent stage may comprise a plurality
of retarder (e.g. birefringent) elements at least two of which have different retardances
(e.g. birefringences), the said plurality of retarder (e.g. birefringent) elements
being combined such that temperature dependent variations in retardance (e.g. birefringence)
of the said plurality of retarder (e.g. birefringent) elements at least substantially
cancel out. The results of such a configuration for apparatus temperatures of 10°C,
20°C and 30°C are shown in Figure 6a, in which quartz and magnesium fluoride (MgF
2) crystals are combined in an athermal combination (in both the birefringent filter
125 and the spectral modulator 115) such that the response of the filter 125 and the
spectral positions of the polarised spectral modulation profiles for each of those
temperatures are substantially the same. This makes for a very stable system, allowing
consistently accurate measurements to be obtained with a calibration plot obtained
at a single temperature as illustrated by Figure 6c which shows plots of fractional
linear polarisation of radiation output by the athermal spectral modulator 115 versus
NO
2 concentration for three different temperatures: 10°C, 20°C and 30°C (which match
each other strongly).
[0164] It has been observed when determining calibration plots that, even when a zero concentration
of the target gas (e.g. NO
2) is observed using the techniques described above, a non-zero value of fractional
linear polarisation may be observed. In this case, the fractional linear polarisation
determined by the controller 120 from which the gas concentration is determined, and
also the calibration reference, may be adapted for a zero target gas concentration
reference. That is, during calibration, a measurement of Q
meas/I
meas and U
meas/I
meas is made for a zero concentration of the target gas (e.g. NO
2) in the calibration volume. These values of Q
meas/I
meas and U
meas/I
meas are then stored to provide zero target gas concentration reference values, Q
meas/I
meas_ref and U
meas/I
meas_ref. These reference values are then taken into account when determining the fractional
linear polarisation of radiation output by the spectral modulator 115, both when calculating
the fractional linear polarisation values for the calibration curve and when calculating
fractional linear polarisation values from which the presence of gas can be determined
in the gas measurement volume. This may be done as follows. Q
meas/I
meas and U
meas/I
meas are determined as before. However, before rotation of the [Q, U] co-ordinate system,
Q
meas/I
meas_ref and U
meas/I
meas_ref are first of all subtracted from Q
meas/I
meas and U
meas/I
meas respectively. Then, when the co-ordinate system is rotated, Q
meas'/I
meas will be indicative of the fractional linear polarisation of the radiation output
by the spectral modulator 115 with respect to the zero target gas concentration reference.
This value may then be compared to the similarly determined calibration curve to determine
the concentration of the gas in the gas measurement volume. Thus, in this case, the
fractional linear polarisation data may comprise a measure of the fractional linear
polarisation of the radiation output by the spectral modulator with respect to a zero
target gas concentration reference.
[0165] The values of fractional linear polarisation in Figure 6c are values of fractional
linear polarisation of the radiation output by the spectral modulator 115 with respect
to the zero target gas concentration reference determined in this way. As the apparatus
used to determine the plots of Figure 6c uses athermal birefringent crystal combinations,
the zero target gas concentration reference need only be determined at a single reference
temperature. This reference can then be used in subsequent determinations of fractional
linear polarisation regardless of the temperature of the apparatus.
[0166] As an alternative to using such athermal combinations of birefringent elements, all
of the retarder (e.g. birefringent) elements of the filter 125 and the spectral modulator
115 may be made from the same crystal material, such as quartz. In this case, the
responses of each of the retarder elements vary in the same way with temperature.
Thus, the response 127 of the birefringent filter 125 and the polarised spectral modulation
profiles shift in frequency by substantially the same amount (and with the same polarity)
as a function of temperature. This is shown in Figure 6b which shows the response
of the birefringent filter 125 and the polarised spectral modulation profiles spectrally
shifting by the substantially same amount in frequency as a function of temperature
from 10°C to 30°C, the birefringent elements of the birefringent filter 125 and spectral
modulator 115 all being made from quartz. As a result, and because there are deep
minima in the response of the birefringent filter 125 at the lower and upper cut-off
frequencies of the bandpass filter 121, the intensity measurements made by the detectors
are not significantly affected by fluctuations in temperature of the apparatus, allowing
accurate measurements to be obtained.
[0167] Additionally or alternatively, the effects of temperature may be compensated for
by measuring the temperature of the apparatus and taking the measured temperature
into account when determining the fractional linear polarisation from which presence
of the gas (and typically gas concentration) is determined. For an apparatus which
does not use an athermal birefringent crystal arrangement, it may be that the zero
target gas concentration reference changes as a function of temperature. It may be
that different zero target gas concentration references are determined for different
temperatures of the apparatus during a calibration phase. In this case, when measurements
of Q
meas/I
meas and U
meas/I
meas are made, it may be that the temperature of the apparatus is measured (either directly
or by calculating the AoLP which varies with temperature) and the appropriate zero
target gas concentration reference is selected in dependence on the measured temperature.
The appropriate zero target gas concentration reference may then be taken into account
to determine fractional linear polarisation of the detected radiation as above. Temperature
of the apparatus can thus be taken into account to thereby more accurately determine
the concentration of the gas. In this case, the fractional linear polarisation data
may comprise a measure of the fractional linear polarisation of the radiation output
by the spectral modulator with respect to a temperature dependent zero target gas
concentration reference.
[0168] This is illustrated in the plots of Figure 6d which shows, for the arrangement to
which Figure 6b relates in which the birefringent elements of the birefringent filter
and retarder(s) are each formed from the same material (in this case quartz), plots
of fractional linear polarisation with respect to a temperature dependent zero target
gas concentration reference versus NO
2 gas concentration for measured temperatures of 10°C, 20°C and 30°C. As shown in Figure
6d, these plots overlap with each other almost completely, indicating that temperature
has been accurately accounted for. For illustration, the fractional linear polarisations
determined at 10°C and 30°C with respect to a zero target gas concentration reference
at 20°C are shown at 698 and 699 respectively in Figure 6d. The deviation of the plots
698, 699 from the other plots of Figure 6d illustrate the improvements in accuracy
which can be achieved by employing a temperature dependent zero target gas concentration
reference.
[0169] Referring back to Figure 1a, it may be that the detection apparatus 112 comprises
a plurality of groups 152, 153 of radiation detectors, each of the said groups comprising
a respective pair of radiation detectors 142, 144 to detect radiation output from
the one or more retarders 108 filtered by respective linear polarisers 146, 148, such
as linear polarisers of a respective group of linear polarisers, for respective linear
polarisation states, the radiation incident on the detectors of the different groups
corresponding to different parts of a scene. Typically the plurality of groups of
radiation detectors are configured to detect radiation output from the one or more
retarders 108 filtered by respective linear polarisers separately and selectively
at the same time as each other. The plurality of groups of radiation detectors may
be provided in a common (e.g. two dimensional) array of radiation detectors. In this
example, each of the groups 152, 153 is intended to detect radiation filtered for
first and second linear polarisation angles of 0° and 90° to thereby make first and
second polarisation dependent intensity measurements from which the presence of the
gas can be determined as set out above in respect of Figure 1a. Thus each of the said
groups 152, 153 selectively detect polarised radiation conforming to each of first
and second of the said polarised spectral modulation profiles (relating to linear
polarisation angles of 0° and 90° respectively) to thereby provide respective first
and second polarisation-dependent radiation intensity measurements from which the
presence of the gas can be determined. As shown in Figure 1, the radiation detectors
of the plurality of groups may be (but are not necessarily) provided on a common focal
plane to thereby provide a compact apparatus. It may be that the radiation detectors
of the plurality of groups are provided in a tessellated or pixelated arrangement.
The polarisers may be provided in a tessellated arrangement in a common focal plane,
for example in a common (e.g. two dimensional) array of polarisers, for example provided
over the radiation detectors of the respective groups.
[0170] It may be that the groups of detectors are arranged in a periodic tessellated arrangement,
the periodicity of the tessellated arrangement being in respect of the linear polarisation
angles of radiation which the detectors of the groups are configured to detect. The
periodicity may be in a common focal plane on which the groups of detectors are provided.
It may be that the corresponding groups of polarisers are provided in a periodic tessellated
arrangement, the periodicity being in respect of the linear polarisation angles of
the said groups of polarisers. The periodicity may be in a common focal plane on which
the groups of detectors are provided. It may be that the controller 120 is to determine,
for each of the said groups, the presence of the gas (preferably the absorption strength
of the gas) and, in dependence on the said determinations, to determine a (typically
two-dimensional) image selectively indicating presence of gas (preferably relative
absorption strengths of the gas) at a plurality of locations across the image. It
may be that the controller 120 is configured to output the said two-dimensional image
to a user interface such as a monitor screen.
[0171] It may be that the said plurality of groups of radiation detectors comprises a first
group arranged to receive radiation incident on the objective lens 102 at a first
angle of incidence and a second group arranged to receive radiation incident on the
objective lens 102 at a second angle of incidence different from the first angle of
incidence. It may be that the first and second angles of incidence are offset from
each other by at least 45°. By detecting received radiation at different angles of
incidence by the apparatus 100, multi-axis differential optical absorption spectroscopy
(MAX DOAS) may be performed on the detected signals using a single pixelated detector
array. It may be that the detector array is tilted to enable detection of radiation
at the required angles of incidence. In this case, it may be that the telecentric
objective lens 102 would need to be replaced by a non-telecentric objective lens.
This is explained in more detail with reference to Figure 7 below.
[0172] Although the groups of Figure 1a each comprise a pair of detectors with a corresponding
pair of linear polarisers, it will be understood that each group may contain more
than two detectors with corresponding linear polarisers. For example, each of the
groups may comprise at least three detectors and corresponding linear polarisers or
at least four detectors and corresponding linear polarisers. For example, each of
the groups may comprise three detectors and three corresponding linear polarisers
of linear polarisation angle 0°, 60° and 120° such that the detectors of each group
detect polarised radiation conforming to the respective polarised spectral modulation
profiles associated with those linear polarisation angles to thereby provide three
polarisation dependent intensity measurements from which presence (and typically concentration)
of the gas can be determined as described above.
[0173] In another example, each of the groups may comprise four detectors and four corresponding
linear polarisers of linear polarisation angle 0°, 45°, 90° and 135° such that the
detectors of each group detect polarised radiation conforming to the respective polarised
spectral modulation profiles associated with those linear polarisation angles to thereby
provide four polarisation dependent intensity measurements from which presence (and
typically concentration) of the gas can be determined as described above. Figure 5a
schematically illustrates an example micropolariser array which may comprise a plurality
of linear polarisers 512n. The micropolariser array can be provided over a detector
array such that there is a 1:1 correspondence between linear polarisers 512n and detectors
of the array. As shown in Figure 5b, the plurality of detectors may comprise a plurality
of groups 152, 153 of radiation detectors and as shown in both Figures 5a and 5b the
plurality of micropolarisers may comprise a plurality of groups of linear polarisers.
502 and 504 illustrate examples of groups of micropolarisers of the micropolariser
array in Figure 5a, however the micropolariser array is not limited to comprise only
two groups of linear polarisers and may alternatively comprise one group or more than
two groups. Each of the plurality of groups may comprise a first 522, second 524,
third 526 and fourth 528 linear polarisers. The respective micro-polarisers may have
linear polarisation angles of 0°, 45°, 90° and 135° respectively, thereby causing
the respective detectors with which they are associated to detect radiation conforming
to polarised spectral modulation profiles of relative phases 0°, 90°, 180° and 270°.
The detected 0°, 45°, 90° and 135° linearly polarised radiation by each group of radiation
detectors over which the groups 502, 504 of micropolarisers are provided corresponds
to first, second, third and fourth polarisation-dependent radiation intensity measurements
respectively from which presence of the gas, absorption strength of the gas and gas
concentration can be determined (see above).
[0174] As shown in Figure 5b, the groups 502, 504 of linear polarisers 512n and the corresponding
groups 152, 153 of detectors may be provided on a common focal plane. The groups of
linear polarisers and the corresponding groups of radiation detectors may be provided
adjacent to each other in a tessellated arrangement to form a repeating 2x2 arrangement
of radiation detectors. For each of the groups of radiation detectors, the presence
of a gas may be determined by a controller 120 processing the first to fourth polarisation
dependent intensity measurements as discussed below. For each group 152, 153 of detectors
512n, there may be first, second, third and fourth polarisation-dependent radiation
intensity measurements and the controller 120 may determine the fractional linear
polarisation to determine an absorption strength of the gas (e.g. NO
2). The controller may determine, using the measurements from each of the groups of
the radiation detectors, a two-dimensional image of the relative absorption strength
of the gas at a plurality of locations across the image. The controller may output
the two-dimensional image to a user interface (e.g. monitor screen).
[0175] By determining the fractional linear polarisation from four polarisation-dependent
intensity measurements, a 2x2 micropolariser array pattern can be provided on the
detectors. Such a tessellation pattern is easier to implement over a relatively large
pixelated array of detectors than, for example, a 2x1 pattern. In other embodiments,
repeating 2x3 or 2x4 micropolariser array patterns may be provided over a pixelated
detector to provide groups of eight polarisation dependent intensity measurements
from which the presence of the gas can be determined. In this case, the angles of
linear polarisation of the respective micropolarisers of each 2x4 array pattern may
be offset from each other by 22.5° so that the corresponding polarised spectral modulation
profiles are offset from each other by 45°. Again, this provides an efficient tessellation
pattern for a large pixelated detector array.
[0176] By a plurality of groups of linear polarisers and corresponding radiation detectors,
periodically arranged in respect of the linear polarisation angles of radiation which
the detectors of the groups are configured to detect or in respect of the linear polarisation
angles of the said groups of polarisers, detecting radiation filtered for particular
polarisation states at different points in space in accordance with a specific pattern,
an additional spatial modulation of the received radiation is provided which enables
the polarisation pattern, and therefore the gas detection or gas concentration pattern,
to be measured across a focal plane. As well as providing information about the gas
content across an area of space, this additional spatial modulation can also help
to increase spatial resolution and remove or reduce systematic errors, for example
from the polarisation-dependent radiation intensity measurements or in fractional
polarisation measurements from which the presence of gas may be determined. In this
case, the polarisation modulation provided by the one or more retarders may provide
the received radiation with polarised spectral modulation profiles which allow the
presence of the gas can be determined, while the additional spatial modulation helps
to improve the accuracy with which radiation conforming to those profiles can be measured.
Making measurements by the plurality of groups also allows an image to be determined
of the relative absorption strength of the gas at a plurality of locations across
the image.
[0177] It may be that the presence (and typically absorption strength or concentration)
of the gas is determined once per group (or repeating pattern) of detectors. However,
in order to improve the spatial resolution of the image and to reduce or avoid aliasing,
it may be that the controller 120 is programmed to determine whether gas is present
for each detector of the array. In order to do this, the controller 120 takes into
account signals detected by proximal (e.g. neighbouring) detectors which conform to
different spectral modulation profiles from the pixel in question to enable the fractional
linear polarisation to be estimated for each pixel as above. The controller 120 may
be configured to use a polarisation-dependent radiation intensity measurement from
one detector in the determination of the presence of the gas for a plurality of (e.g.
two or more, three or more or four or more) positions in the focal plane. The controller
120 may be configured to use a polarisation-dependent radiation intensity measurement
from each of a plurality of detectors in the determination of the presence of the
gas for a respective plurality of positions in the focal plane. This may be facilitated
by the additional spatial modulation provided by the periodic tessellated arrangement
of the detectors, the periodicity of the tessellated arrangement being in respect
of the linear polarisation angles of radiation which the detectors of the groups are
configured to detect. For example, the controller 120 may be configured to determine
whether gas is present in respect of a first position on the focal plane depending
on polarisation-dependent radiation intensity measurements by at least a first one
of the radiation detectors and a second one of the radiation detectors, and to determine
whether gas is present in respect of a second position on the focal plane depending
on polarisation-dependent radiation intensity measurements by at least the first one
of the radiation detectors and a third one of the radiation detectors different from
the second one of the radiation detectors. It may be that the controller 120 is to
determine the presence of the gas at a first position in dependence on polarisation-dependent
radiation intensity measurements by at least a first detector, a second detector and
a third detector and at a second position in dependence on polarisation-dependent
radiation intensity measurements by at least the first detector, the second detector
and a fourth detector different from the third.
[0178] As well as permitting an increase in the spatial resolution of the image and reducing
or avoiding aliasing, this technique can also reduce systematic errors, particularly
if the plurality of groups of radiation detectors are periodically arranged across
the focal plane in respect of the linear polarisation angles of radiation which the
detectors of the groups are configured to detect.
[0179] It may be assumed that the gas concentration varies smoothly, and therefore that
the fractional linear polarisation varies smoothly (e.g. at a frequency greater than
the individual pixel level or at a frequency greater than the spatial modulation frequency)
across the radiation detectors of the plurality of groups (e.g. over the focal plane
of the detector array). As such, a systematic issue with an individual pixel can be
mitigated by the spatial modulation. In the Fourier domain, this can be thought of
as considering information at a predefined spatial modulation carrier frequency (i.e.
the frequency with which the polarisation is spatially modulated across the detectors)
and removing or discarding information at frequencies which differ (or which differ
significantly) from the predefined spatial modulation carrier frequency (which may
be information from individual pixels producing systematic errors).
[0180] As an alternative way to increase the spatial resolution of the image, fractional
linear polarisations estimated for each group or repeating pattern can be interpolated
to provide estimated fractional linear polarisations for each detector of the array.
[0181] As illustrated in Figure 4a, an alternative detection apparatus 112 may be provided
which comprises an adjustable polariser 112a and a radiation detector 112b. The adjustable
polariser 112a may be used to filter the radiation sequentially at different linear
polarisation angles before it is detected by one or more co-planar detectors provided
at a detector plane 112b. The adjustable polariser 112a may be a rotatable polariser
which rotates to sequentially transmit different polarisations of radiation. This
may result in the polarisation of the radiation by the one or more detectors varying
with time. The polarisation dependent intensity measurements are made by the detection
apparatus 112 in this case by filtering the radiation incident on the detection apparatus
at different linear polarisation angles by virtue of the adjustable polariser 112a
transmitting radiation of different linear polarisations to the detectors at the detector
plane 112b over time. For example, in order to implement a similar example to the
first example described above in which two polarisation dependent intensity measurements
are made by filtering radiation from the spectral modulator 115 at two different angles
of linear polarisation using two separate fixed angle polarisers and two detectors,
the adjustable polariser 112a may be provided in a first rotational orientation to
transmit only linearly polarised radiation at 0°for a first time period. Within this
first time period the linearly polarised radiation at 0° is incident on the detectors
at the detector plane 112b and one or more detectors at the detector plane 112b may
provide a first polarisation dependent intensity measurement. The linear polarisation
angle of the adjustable polariser 112a may then be rotated so that the adjustable
polariser 112a is rotationally orientated to transmit radiation linearly polarised
at 90°. For a second time period the 90° linearly polarised radiation is incident
on the one or more detectors at the detector plane 112b and one or more detectors
at the detector plane 112b may provide the second polarisation dependent intensity
measurement. Based on the first and second polarisation-dependent intensity signal
measurements presence or absorption strength or concentration of the gas can be determined
as before. The adjustable polariser 112a is not limited to only transmit linearly
polarised radiation at 0° and 90°, and may be used to transmit radiation linearly
polarised at any angle, allowing the detection apparatus 112 of Figure 4a to be used
as an alternative to separate fixed angle linear polarisers in any of the embodiments
described above.
[0182] As an alternative to providing a rotatable polariser 112a, as shown in Figure 4b
a rotatable (typically rotating) half wave plate 112c may be used in combination with
a fixed linear polariser in front of the detector 112b, to perform the same purpose
as the adjustable polariser 112a to allow sequential detection of radiation at different
polarisation angles by one or more detectors.
[0183] Temporally modulating the polarisation of radiation detected by the detector plane
112b (rather than detecting radiation filtered for different polarisations at the
same time) may have the advantage that the polarisation-dependent intensity measurements
may be independent from specific transmissions and gains from different individual
radiation detectors on the detection apparatus, or individual pixels. However, measurements
taken in sequence rather than in parallel may be susceptible to time variations in
the scene which is being imaged and in the illumination.
[0184] The detector plane 112b may be any suitable radiation detector which may measure
intensity of linearly polarised radiation at any polarisation angle and can distinguish
signals measured over a different time frame.
[0185] As well as providing an alternative to fixed angle linear polarisers provided over
radiation detectors as discussed above, modulating the polarisation state of radiation
transmitted from the spectral modulator 115 to the detector(s) over time can be used
to provide an additional stage of modulation to fixed angle polarisers provided over
respective detectors which can help remove systematic errors in the determination
of the presence of the gas, and from gas concentration measurements. This is because,
by modulating the polarisation state of radiation transmitted from the spectral modulator
115 over time, the component of radiation output by the spectral modulator 115 whose
polarisation angle is modulated by the temporal polarisation modulator to a linear
polarisation angle which is transmitted by the linear polariser(s) provided over the
detector(s) changes over time. Accordingly, the same detector(s) 112, 112b can be
made to detect radiation output by the spectral modulator 115 at different angles
of linear polarisation over time. Thus, the same detector(s) 112, 112b can be made
to detect radiation output by the spectral modulator 115 conforming to different polarised
spectral modulation profiles over time. This helps to remove or at least reduce the
said systematic errors. For example, it may be that for a first time period the polarisation
state of radiation transmitted from the spectral modulator 115 to the detector(s)
may cause a first radiation detector to detect radiation filtered for a first linear
polarisation angle and a second radiation detector to detect radiation filtered for
a second linear polarisation angle different from the first linear polarisation angle
in order to provide first and second polarisation dependent intensity measurements.
These measurements may then be used to determine fractional linear polarisation, and
thus the presence of the gas. In order to determine whether the first or second detectors
are responsible for any systematic errors, the polarisation state of radiation transmitted
from the spectral modulator 115 to the detectors may be modulated differently such
that for a second time period the first detector is caused to detect radiation filtered
for the second linear polarisation angle and the second radiation detector to detect
radiation filtered for the first linear polarisation angle in order to provide third
and fourth polarisation dependent intensity measurements. These measurements may then
be used to determine fractional linear polarisation, and thus the presence of the
gas. As discussed above, the determination of the fractional linear polarisations
may in each case involve a subtraction operation between the first and second, and
fourth and third, polarisation dependent intensity measurements. As the first and
fourth and second and third polarisation dependent intensity measurements should be
substantially the same as each other, the outputs of the respective subtraction operations
should be substantially the same as each other. However, if either of the detectors
is responsible for causing a systematic error, this may not be the case. Thus, a systematic
error caused by either detector can be identified. The controller 120 may then readily
remove or reduce the systematic error, for example by omitting information from the
detector causing the error in the determination of fractional linear polarisation.
[0186] Additionally or alternatively, a periodic temporal modulation of the polarisation
output from the spectral modulator 115 to the detectors can provide additional information
similar to that of the spatial modulation discussed herein by way of which systematic
errors can be reduced.
[0187] As mentioned above, Figure 7 illustrates a modification of the apparatus 100 in which
the telecentric objective lens 102 is replaced by a non-telecentric lens 702. Radiation
received and output by the non-telecentric lens 702 may be incident on filter 104,
and hence the spectral modulator 115, at an angle of incidence which is dependent
on its angle of incidence on the non-telecentric lens 702. The spectral modulator
115 acts as previously described to provide radiation having a polarisation state
spectrally modulated in accordance with a plurality of polarised spectral modulation
profiles, each associated with a different linear polarisation angle. In addition,
the birefringence of the one or more retarders 108 is dependent on the angle of incidence
of radiation. This is described by the following formula for calculating the phase
delay of a birefringent crystal for radiation of wavelength λ:

where ΔΦ is the phase delay, θ is the angle of incidence, n
e is the refractive index of the extraordinary axis of the crystal (which is angle
of incidence dependent) and n
O is the refractive index of the ordinary axis of the crystal.
[0188] In addition, the path length through the retarders 108 varies with angle of incidence.
[0189] This provides a further (angular) stage of modulation on the polarisation of radiation
such that the polarisation of radiation output from the spectral modulator 115 filtered
for the same linear polarisation angle, but which was incident on the one or more
retarders 108 at different angles of incidence, is spectrally modulated in accordance
with respective (different) polarised spectral modulation profiles associated with
the same linear polarisation angle but which are offset in phase from each other.
Similarly to the additional spatial modulation, this allows additional information
to be obtained, which helps to remove or at least reduce systematic errors.
[0190] As before, radiation detected at linear polarisation angles associated with the polarised
spectral modulation profiles may have an intensity which varies in accordance with
the respective polarised spectral modulation profile. Such polarisation filtered radiation
may be detected to provide polarisation-dependent radiation intensity measurements
at the radiation detection apparatus 112. Accordingly, in the present embodiment,
two or more of the first, second (and third and/or fourth and so on) polarisation
dependent intensity measurements from which the gas can be detected may be radiation
filtered for the same angle of linear polarisation but incident on the one or more
retarders 108 at different angles of incidence. This is illustrated in Figure 7 by
fixed polarisers 746, 747 over first and second detectors 142, 144 having the same
linear polarisation angle of 0°. From the polarisation-dependent intensity signal
measurements the fractional linear polarisation may be calculated using the above
equations as previously described. This fractional linear polarisation may then be
converted into an absorption strength of the gas using a calibration curve or any
other suitable method. As before, the more distributed the phases of the polarisation
filtered radiation which is detected by the detector throughout a 360° cycle, the
more efficient the fractional linear polarisation measurement will be and the more
sensitively gas can be detected.
[0191] In some embodiments, polarisation dependent intensity measurements may be obtained
from radiation filtered for each of a plurality of linear polarisation angles, but
for each linear polarisation angle multiple polarisation dependent intensity measurements
may be obtained of radiation incident on the objective lens 702 (and thus the one
or more retarders 108) at different angles of incidence.
[0192] Thus, it may be that the spectral modulator is to spectrally modulate the linear
polarisation state of the received radiation to provide the received radiation with
a plurality of polarised spectral modulation profiles which are offset in phase from
each other, each of which is associated with a different linear polarisation angle,
and that the apparatus is to additionally modulate the polarisation of the received
radiation in one or more other domains, such as the angular, temporal and/or spatial
domains. In this case, the polarised spectral modulation profiles to which the radiation
detected by the detectors conform may be provided by the said spectral modulation,
while the additional modulation helps to improve the accuracy with which radiation
conforming to those profiles can be measured. Modulating polarisation of radiation
in more than one modulation domain (e.g. angle of incidence, frequency, spatial, time)
helps to remove systematic noise/interference from the measurements. For example,
there may be large systematic errors for individual groups of pixels (e.g. if there
are dead pixels in particular groups). When an additional (e.g. angular, spatial or
temporal polarisation) stage of modulation is applied over the field of view, systematic
differences for all pixels with the same polarisation filtering can be readily removed
or at least reduced. In addition, if there is a priori knowledge about the scene being
imaged, e.g. that the signal should vary smoothly over the field of view, this can
be used to correct for local pixel effects. Thus, it may be that the controller 120
is configured to remove or reduce systematic errors from the gas detection measurements
depending on a combination of polarisation modulations in different domains. For example,
the controller may be configured to reduce or remove systematic errors from the polarisation-dependent
intensity measurements, or fractional polarisation measurements from which the presence
of gas may be determined, depending on the said additional modulation.
[0193] As mentioned above, an additional spatial modulation provided by a plurality of groups
of radiation detectors detecting radiation output from the one or more retarders filtered
for respective linear polarisation states, which may vary periodically across the
focal plane, can be used by the controller 120 to remove or reduce systematic errors
in the gas detection measurements. By detecting the received radiation, and thus the
presence of gas, across the focal plane with uniform or smoothly varying polarisation,
additional information is provided as compared to merely combining pixel values from
a particular group to determine the presence of the gas, of which any two may be erroneous.
As discussed above, additional information can additionally or alternatively be provided
by an additional polarisation modulation in the angular or time domains which allows
systematic errors to be mitigated by the controller 120.
[0194] In all of the above described embodiments, the fractional linear polarisation may
be relatively small (~10
-3) and hence steps may be taken to mitigate systematic effects that may result in erroneous
measurements. As discussed, additional spatial, angular or temporal stages of modulation
may be added to the apparatus to reduce systematic effects. The additional stages
of modulation are not essential to the present disclosure and may only enhance the
performance of the apparatus.
[0195] Any of the above embodiments may be used in a remote atmospheric trace gas monitoring
system for capturing radiation which has passed through the atmosphere at different
angles and comparing them.
[0196] Figure 8 is a flow chart schematically illustrating a method for determining the
presence of a gas using an apparatus as described in any one of the above embodiments.
[0197] At 902, received radiation within a selected frequency range may be selectively transmitted,
e.g. by a filter. An absorption spectrum of a gas may have a plurality of lines offset
in frequency from each other within the selected frequency range. The selective transmission
may be performed using a filter with a suitable transmission spectrum, for example
a bandpass filter 121 optionally in combination with one or more birefringent filter
stages 125. At 904, the polarisation of received radiation may be spectrally modulated
within the selected frequency range. The polarisation of received radiation may be
spectrally modulated in accordance with a plurality of polarised spectral modulation
profiles which may be offset in phase from each other. The spectral modulation may
be performed by one or more retarders.
[0198] At 906, at least a portion of the spectrally modulated radiation may be selectively
detected which may be filtered for a respective polarisation state. The detected radiation
may conform to each of at least first and second of the said polarised spectral modulation
profiles.
[0199] At 908, at least first and second polarisation-dependent radiation intensity measurements
may be provided that may correspond to each of at least first and second of the said
polarised spectral modulation profiles respectively. The at least first and second
polarisation-dependent radiation intensity measurements may be detected using a radiation
detection apparatus 112.
[0200] At 910, the presence of a gas may be determined based on the at least first and second
polarisation dependent radiation intensity measurements. This may comprise determining
the fractional linear polarisation based on the at least first and second polarisation
dependent radiation intensity measurements. The method may further include determining
a relative absorption strength of the gas. The method may further include determining
a concentration of the gas by comparing the determined fractional linear polarisation
to one or more calibration references.
[0201] It will be understood that any number of polarisation dependent intensity measurements
(preferably any number greater than one) may be made from which the presence of the
gas can be determined, each of the polarisation dependent intensity measurements corresponding
to a respective polarised spectral modulation profile of the plurality of polarised
spectral modulation profiles in accordance with which the spectral modulator 115 modulates
the radiation.
[0202] Although the linear polarisers described above each receive from the one or more
retarders radiation having a plurality of linear polarisation components and having
a single output to provide radiation having a single predefined linear polarisation
to a respective detector, it will be understood that in some embodiments polarising
beam splitters may additionally or alternatively be provided, the polarising beam
splitters each receiving from the one or more retarders radiation having a plurality
of linear polarisation components and providing at separate outputs polarised radiation
at different predefined (e.g. orthogonal) linear polarisations to respective detectors.
[0203] For the purposes of the description, a phrase in the form "A / B" or in the form
"A and/or B" means (A), (B), or (A and B). For the purposes of the description, a
phrase in the form "at least one of A, B, and C" means (A), (B), (C), (A and B), (A
and C), (B and C), or (A, B and C).