Test impacts of cloud and aerosols on methane retrievals using 3D CTM.
Qian
Tan, Dept. Earth, Atmospheric, and Planetary Sci., Massachusetts Institute of Technology, qiantan@mit.edu
(Presenting)
Ronald
Prinn, Dept. Earth, Atmospheric, and Planetary Sci., Massachusetts Institute of Technology, rprinn@mit.edu
Christian
Frankenberg, Institut fuer Umweltphysik, University of Heidelberg, Christian.Frankenberg@iup.uni-heidelberg
Yu-Han
Chen, Dept. Earth, Atmospheric, and Planetary Sci., Massachusetts Institute of Technology, yuhanchen10@hotmail.com
We compared our model (MATCH driven by NCEP Reanalysis data) simulations of methane column burdens with remotely sensed data from SCIAMACHY for the year 2003. The model agrees very well with observations in terms of spatial distribution. The linear regression slope is 1.01 for yearly averages, and 0.96-1.05 for monthly averages. The model has 2.7E18 molecules/cm2 offset (~8% of the total CH4 column) compared with observations, when using effective cloud heights estimated by the retrieved CO2 column. In order to remove this offset, we further investigated the impact of cloud and aerosols on methane column estimates. The correlation of SCIAMACHY methane column with the cloud top height, estimated using the FRESCO scheme, increases with cloud fraction. When the sky is overcast, they are linearly correlated with an r2 of ~0.8. When the FRESCO cloud information, including cloud fraction and effective cloud height, is incorporated, we found that the model-satellite comparison splits into two branches: one is correlated with the cloud top height, while the other is not, especially when the cloud fraction is small. We then tested the impact of aerosols on the methane column, cloud fraction and cloud top height retrievals. We used MODIS aerosol optical thickness (AOT) at 550 nm, when it is overlapped with SCIAMACHY orbits. MODIS and SCIAMACHY footprints have about 30-60 min temporal differences. We assume the aerosol distribution does not vary during this time period.