Variational Retrieval of Cloud Parameters from GOES Sounder Longwave Cloudy Radiance Measurements

The optimal nonlinear inversion or one-dimensional variational (1DVAR) method was used to retrieve the cloud-top height and effective cloud amount from Geostationary Operational Environmental Satellite (GOES) sounder longwave spectral-band cloudy radiance measurements. The cloud-top pressure and effective cloud amount derived from the carbon dioxide (CO2)–slicing technique served as the background or first guess in the 1DVAR retrieval process. The atmospheric temperature profile, moisture profile, and surface skin temperature from the forecast analysis were used for the radiative transfer calculation in both the CO2-slicing method and the 1DVAR retrieval processing. Simulation studies were made to investigate the accuracy (the retrievals were compared with truth) of the cloud-top pressures and the effective cloud amounts derived from both the CO2-slicing and 1DVAR algorithms. Significant improvement of 1DVAR over CO2-slicing cloud properties was found in the simulation studies; an improvement of 10–50 hPa for root-mean-square error was obtained in 1DVAR over the CO2-slicing-derived cloud-top pressures, depending on the cloud height (high, mid, or low). This improvement came largely from the reduction of the bias in the 1DVAR retrievals over the CO2-slicing cloud-top pressures. The 1DVAR approach was applied to process the GOES-8 sounder cloudy radiance measurements; consistent with the simulation results, CO2 slicing assigned high and low clouds to lower levels than 1DVAR did.