 |
| Titel |
A retrieval algorithm for approaching XCH4 from satellite measurements: Sensitivity study and preliminary test |
| VerfasserIn |
Jianbo Deng, Yi Liu, Dongxu Yang, Zhaonan Cai |
| Konferenz |
EGU General Assembly 2014
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 16 (2014) |
| Datensatznummer |
250094325
|
| Publikation (Nr.) |
 EGU/EGU2014-9730.pdf |
|
|
|
|
|
| Zusammenfassung |
| Satellite measurements of column-averaged dry air mole fractions of CH4 (XCH4) in
shortwave infrared (SWIR) with very high spectral resolution and high sensitivity near the
surface, such as the Thermal And Near-infrared Sensor for carbon Observation (TANSO)
onboard the Green gas Observing SATellite (GOSAT, launched 2009), are expected to
provide the large spatial and temporal information on the sources and sinks of CH4, which
would contribute to the understanding of CH4 variation in global region and its impact on
climate change. One of the important science requirements of monitoring CH4 from
hypsespectral measurements is to establish a highly accurate retrieval algorithm. To approach
XCH4efficiently, we developed a SWIR two-band (5900-6150 cm-1 and 4800-4900
cm-1) physical retrieval algorithm after a series of sensitivity study. The forward
model in this algorithm was based on a vector linearized discrete ordinate radiative
transfer (VLIDORT) model coupled with a line-by-line radiative transfer model
(LBLRTM), which was applied to realize online calculation of absorption coefficient
and backscattered solar radiance. The information content of CH4, H2O, CO2 and
temperature in different retrieval band and bands combination was investigated
in order to improve the algorithm. The selected retrieval bands retains more than
90% of the information content of CH4, CO2, and temperature, and more than
85% of that of H2O. The sensitivity studies demonstrate that the uncertainty of
H2O, temperature and CO2 will cause unacceptable errors if they were ignored, for
example, a 10% bias on H2O profile will lead to 50 ppb retrieval error, and a 5 K
shift on temperature profile will cause 20 ppb error to the result while CO2 has
little influence. The simulated retrieval test shows it is more efficient to revise the
influence of temperature and H2O with a profile model than with a temperature
offset and a H2O scale factor model. A preliminarily retrieval test using GOSAT
Level 1B spectral data shows that most of the retrieval error was less than 1%. |
|
|
|
|
|
|