Retrieval & monitoring of atmospheric green house gases (gh gs) through remote sensing

Heavier precipitation,
more intense and longer droughts….

CLIMATE CHANGE
ATMOSPHERIC AEROSOL GREEN HOUSE GASES (GHGs)
GLOBAL WARMINGGLOBAL DIMMING

GLOBAL AVERAGES OF THE CONCENTRATIONS OF CARBON DIOXIDE,
METHANE, NITROUS OXIDE, CFC-12 AND CFC-11
These gases account for about 97% of the direct warming effect of the long-lived
greenhouse gases since 1750. The remaining 3% is contributed by an assortment
of 10 minor halogen gases. ( Source NOAA, Annual Greenhouse Gas Index )
CO2
N2O
CH4 CFCs

Retrieval & Monitoring of AtmosphericRetrieval & Monitoring of Atmospheric
Green House Gases (GHGs) throughGreen House Gases (GHGs) through
remote sensingremote sensing
Debasish Chakraborty
Roll No. – 4843
Division of Agricultural Physics

RETRIEVAL: To find or extract stored information.
MONITORING: To watch & check over a period of time in order
to see how any phenomena develops/changes so that one can take
necessary action.
So, monitoring of Green House Gases(GHGs) over the
globe is a spatiotemporal property.
GHGs Measurement
Conventional
Remote sensing

Standard type of technique to measure GHGs
VIALS
GC analysis
ECD, FID
detectors
Gas storage
Gas accumulation over time
Closed chamber – Gas Chromatographic analysis,
IRGA

Advantage:
Technique is simple
Can be handled with short training
Very accurate
Limitation:
Limited spatial distribution
Sampling error
Closed chamber – Gas Chromatographic analysis,
IRGA

SATELLITE MEASUREMENTS
DVANTAGE:
 Provide Global coverage
 High temporal resolution
 Data with sufficient precision is becoming available
- Multi-purpose missions-SCIAMACHY,AIRS
-Missions dedicated to GHGs-GOSAT/JAXA ; launched on jan,200
LIMITATION:
 Absolute measurement of physical parameters
 Several disturbances (sensor cal/val, clouds, aerosol etc)
 Retrieval needs complex algorithms
 Asks for expertise

Ground Based Project :
FLUXNET
LIMITATION
DISCONTINUITY OF
MEASUREMENTS
COORDINATION BETWEEN
STATIONS
UPSCALING METHODS
LIMITED AREA COVERAGE
SATELLITE MONITORING COMBINED WITH THIS
GROUND BASED PROJECTS CAN BE A BETTER
OPTION

ATMOSPHERIC SCIENCE SPACEBORNE INSTRUMENTS

CURRENTLY WORKING
SATELLITES
MID & THERMAL INFRARED REGION(TIR & MIR)
HIRS(2002) - NOAA
AIRS(2002) - NASA
IASI(2006) - EUMETSAT
DETECTION
Thermal radiation emitted from surface & atmosphere
(3.6 to 15µm)
ADVANTAGE:
Day & night measurement is possible
DISADVANTAGE:
Lack of sensitivity in lower troposphere

CURRENTLY WORKING SATELLITES
UV/VIS/NIR/SWIR REGION
SCIAMACHY(2002) - ESA
TANSO(2009) - JAXA
OCO(2009) - NASA
DETECTION
Reflected, backscattered, transmitted & emitted from
surface & atmosphere (240 to 2400 nm)
DISADVANTAGE:
Restricted to day only
ADVANTAGE:
Sensitivity constant with height
& maximum near the surface

ABSORPTION BANDS OF DIFFERENT CONSTITUENTS

POSSIBLE ERROR SOURCES
EVALUATED IN ADVANCE:
 Spectroscopic parameters
 Solar spectra
CORRECTED BY ADDITIONAL
INFORMATION
 Cloud covered scene
 Aerosol covered scene
 Surface elevation
 Surface spectra
 Water vapor
 Temperature
Cirrus effect can be cancelled by 760 nm(O2
band) and 2000nm (H2O saturated spectral
region).

MEASURED DATA FILTERING FOR NOISE REMOVAL
FILTERING
ITEMS
Solar Zenith
Angle
Cloud
Estimation
Aerosol at high
Altitudes
Filtered
spectra
Aerosol Transport
Model
(ex.SPRINTARS)
Cloud
Input
Spectra

ATMOSPHERE
R T MODEL
INITIAL CONSTITUENTS
TEMPERATURE
PRESSURE
ALBEDO
DATA PROCESSING
SYNTHESIZED
SPECTRA
FILTERED SPECTRA
Yokota et. al

CASE STUDY-I
STUDY AREA: Boreal forests (Novosibirsk region) & the region
of Surgut
SENSOR USED: AIRS
AMSU-A
RADIATIVE TRANSFER MODEL: SARTA
RUSSIAN METEOROLOGY AND HYDROLOGY Vol. 34 No. 4 2009

1. SELECTION OF C02 SENSITIVE CHANNELS
CO2-sensitive channels at low sensitivity to
interfering factors
 Nine LW-channels in the spectral range of 699–
705 cm–1
 Six SW-channels in the spectral range of 1939–
2017 cm–1
THE STUDY HAS TWO PARTS:
∆TB(i) = δTB(i) +δqH2O TB(i) + δqO3TB(i) + δqTB(i) + . . . . + ξi

2. AIRS DATA INVERSION
Analysis of satellite data to sample cloud free measurement or
measurements reduced to cloud cleared conditions
(http://disc.gsfc.nasa/AIRS/data)
Inverse problem in respect to Xco2 is solved numerically using the
Gauss-Newton iteration algorithm, two independent estimates of
Xco2 (LW) & Xco2 (SW) are estimated by AIRS data
Sampling of estimates Xco2(LW) & Xco2(SW) derived for time interval
and the sounding area are subject to spatiotemporal filtering
The results of aircraft CO2measurements (spatially coincident and
quasi- synchronous with satellite )at different altitudes are used for
comparison
The systematic biases is calculated by -
δ(ᾳ) = [TB
obs
(ᾳ) - TB
calc
(ᾳ)], ᾳ= 1, . . . ., n,

The standard deviations (SD) of Xco2(sat) from the aircraft observations
at altitudes 7 and 3 km were calculated to estimate the errors of the
results of the satellite sounding. The SD are 1.5 and 1.2 ppm compared
to the aircraft CO2 observations at altitudes 7 and 3 km, respectively.
Fig: comparison of satellite(2) and aircraft data of 7000 m (1) & 3000 m (3)
RESULT COMPARISON
Novosibirsk
Surgut

CASE STUDY-II
STUDIED GAS: Methane(CH4)
SENSOR USED: SCIAMACHY (Channel 8 – 2260 to 2385 nm )
RADIATIVE TRANSFER MODEL: SCIATRAN
RADIATIVE TRANSFER ALGORITHM: WFM-DOAS
Atmos. Chem. Phys. Discussion., 4,2004

THE WFM-DOAS RETRIEVAL ALGORITHM
Based on fitting a linearised radiative transfer model Ii
mod
plus a low order polynomial Pi to the algorithm of the ratio
of a measured nadir radiance & solar irradiance spectrum,i.e.
observed sun-normalised radiance Ii
obs
.
The WFM-DOAS equation can be written as-
|In Ii
obs
(Vt
) – [ In Ii
mod
(V) + ∑δ Ii
mod
/ δvj /( vj – vj ) + Pi (am )]|2
= |RESi|2
→min
The fit parameters are the desired “trace gas vertical column Vj”
and the polynomial coefficient am”.
Fit parameters are determined by LEAST SQUARE method
PRINCIPLE: Differential detection of radiance in gaseous
absorption channels with respect to neighbouring
atmospheric transparent spectral channels (not influenced
by gas) ,to detect the conce. Of desired gas.

Parameters:
Cloud condition- UV PMD(Polarization Measurement Device)
SCIAMACHY
Standard atmospheric condition-CH4, CO2 current concentration
Tropospheric and stratospheric condition- aerosol
Surface albedo and solar zenith angle
Surface elevation
Water vapour column and temperature profile shift
The reference spectra was generated by-
Radiative transfer model- SCIATRAN

WFM-DOAS CH4 VERTICAL COLUMN RETRIEVAL ERROR(A) USING
SIMULATED MEASUERMENTS
ERROR B-TEMPERATURE PROFILE SHIFT is included
RESULT DISCUSSION

Tab; Comparison of SCIAMACHY WFM-DOAS v 0.5 with
ground based FTS measurement.
N= no. of SCIAMACHY measurements compared with FTS.
Result And Discussion

Fig. Methane column averaged mixing ratios as retrieved from
SCIAMACHY WFM-DOAS V 0.5.
RESULT DISCUSSION

APPLICATION
STUDY AREA: Low and mid latitudes of Northern Hemisphere
SENSOR USED: SCIAMACHY (1558 to 1594 nm)
ALGORITHM USED: WFM-DOAS version 1.0
Atmos. Chem. Phys. Discuss.,7,2007

ALGORITHM: WFM-DOAS version 1.0
 An improved version (Schneising et al., 2007).
 The main problems of the previous version WFMDv0.4 (Buchwitz et
al.,) was solved using spectra with improved calibration
 Better consideration of surface spectral reflectivity variability
 It is no longer required to apply a quite large empirical scaling factor
as was necessary for WFMDv0.4.
QUALITY FILTERING OF SCIAMACHY:
For cloud detection the measured oxygen column(755 to 775 nm) and
PMDs is used.
Ground altitude(pressure) used in simulation by WFM-DOAS increased
above
4.1 km.
To reject ground scenes with strong aerosol contamination, additional
filtration of the SCIAMACHY XCO2 measurements using NASA’s
Absorbing Aerosol Index (AAI) data product from TOMS/ Earthprobe
was done.
Concentration of CO2 can only be retrieved over land , not over sea.

Fig. Atmospheric CO2 over the northern hemisphere
during 2003–2005 as retrieved from SCIAMACHY
satellite measurements.
FIRST DIRECT OBSERVATION OF ATMOSPHERIC CO2 IN
YEAR TO YEAR FROM SPACE

Fig; Satellite retrieved XCO2 and NOAA ESRL Carbon Tracker global
assimilation system data
Increase in the amplitude of the CO2 seasonal cycle with the increase in
latitude
In the retrieved XCO2 seasonal cycle an error of 2ppm is seen.
ERROR & IT’S CORRECTION:

The correction equation is –
DIF=a + b*AMF
Where, DIF=difference between SCIAMACHY &
Carbon Tracker
AMF=1/cos(SZA) + 1/cos(LOS)
where,
AMF=Air mass factor
SZA=Solar zenith angle
LOS=Line of sight scan angle

APPLICATION:
STUDY AREA: 50
N to 67.50
S , 54.50
E to 1470
E
SENSOR USED:
SCIAMACHY (Channel 8 – 2259 to 2361 nm) with WFM-DOAS V 0.4
SPOT-VEGETATION
ISPRS Archives XXXVIII-8/W3 Workshop Proceedings: Impact of climate
change on agriculture,2009

METHODOLOGY
Global weekly ENVISAT-SCIAMACHY
CH4 conce. (ppbv) data of 2004 &
2005
Global 10 days composite of
SPOT-NDVI products of 2004 &
2005
Computed mean monthly CH4 (ppbv)
Study area was extracted by overlying the region’s boundary
and gridded to 0.50
x 0.50
latitude/longitude grid
Validation using NOAA-
CMDL Global view data
Spatial and temporal
variability over study area
CH4 data covering Kharif
season (May-Oct) for
2004 & 2005
NDVI data covering Kharif
season(May-Oct) for 2004 &
2005
Correlation between CH4 conce. & NDVI during
Kharif season over study area
Computed mean monthly NDVI

Fig 1. Temporal and Spatial
Variation of Atmospheric CH4
Concentration Over India
During 2004 – 05
RESULTS:
Fig2. Temporal Variation of
Vegetation Over India
During 2004–05

ig 3. Two year kharif season averaged CH4 conce.
Fig 4. Two year kharif season averaged NDVI.
Fig 5. Correlation between CH4 Conc. and
Vegetation During Kharif Season in 2004-05

PROBLEM OF THESE STUDIES:
Scattering at aerosol and/or cloud particles remains a major
source of uncertainty for SCIAMACHY XCO2 retrievals
The XCO2 retrieval error may amount to 10% in the
presence of mineral dust aerosols.
Houweling et al. (2005)
The thin scattering layer with an optical thickness of 0.03
in the upper troposphere can introduce XCO2 uncertainties
of up to several percent. Schneising
et al. (2008)
Unfortunately, thin clouds with optical thicknesses below 0.1
cannot easily be detected within nadir measurements in the
visible and near infrared spectral region.
Reuter et al., 2009; Rodriguez et al.
(2007).

Algorithm: Merged fit windows approach.
Radiative transfer model: SCIATRAN 3.0
Atmos. Meas. Tech., 3, 209–232, 2010

The measurement vector y consists of SCIAMACHY
sun-normalized radiances of two merged fit windows
concatenating the measurements in the CO2 and O2 fit
window.
= ( , ) +y F x b ԑ
X = state vector
b = parameter vector
&, =ԑ error
The information about these parameters comes mainly from the
O2 measurements and is made available in the CO2 band by the merged
fit windows approach..

The accuracy for scenes with optically thin cirrus clouds was
drastically enhanced compared to a WFM-DOAS like retrieval.
At solar zenith angles of 400
, the presence of ice clouds with
optical thicknesses in the range of 0.01 to 1.00 contributed with
less than 0.5 ppm to the systematic absolute XCO2 error if a
perfect forward model is assumed.
RESULTS:

Conclusions:
 Green House Gases (GHGs) can be measured with good
accuracy from satellite data if proper algorithms are
applied
 Through inverse modeling of measured GHGs we can
know in detail about their sources and sinks
Further development in understanding about different
factors, their interactions influencing the GHG retrieval
and improvement in mathematical methods will surely be
able to predict GHGs with better accuracy
Monitoring of GHGs emitted from agricultural practices
and activities, wetlands over a region can be done with
good accuracy

Retrieval & monitoring of atmospheric green house gases (gh gs) through remote sensing

More Related Content

What's hot (20)

Viewers also liked (11)

Similar to Retrieval & monitoring of atmospheric green house gases (gh gs) through remote sensing (20)

Recently uploaded (20)

Retrieval & monitoring of atmospheric green house gases (gh gs) through remote sensing

Editor's Notes