Cloud Top Pressure from a Subset of AIRS Channels

The retrieval of cloud top pressure (CTP) and effective cloud fraction (ECF) from the NASA EOS AIRS instrument via the CO2 slicing technique is investigated and compared to similar approaches applied to GOES and MODIS. Several subsets of the AIRS channels in the 12-14 micrometer region which minimized redundancy but provide sensitivity to clouds throughout the atmosphere are used in the retrieval technique and compared to the four available CO2 channels of the other instruments. Simulated data was used to understand retrieval sensitivity, biases between the channel subsets, and limitations of the approach. The observed data was the basis for a detailed inter-comparison between the datasets and validation against ground truth lidar data. The results indicate that retrieval bias varied from 20-50hPa with the larger values occurring at lower pressures. The AIRS channel subsets produced more successful CTP retrievals for clouds with low ECF values than did GOES, particularly for low clouds. The CO2 slicing algorithm became unstable and produced large retrieval errors in the presence of strong tropospheric inversions. On the edges of cloud systems or for low ECFs, the AIRS provided more realistic retrievals than GOES by maintaining spatial continuity in the cloud patterns in these regions. Limited comparisons with micropulse lidar data from the ARM/CART site confirmed the improved retrieval accuracy over GOES.