The lack of a global characterization of upper tropospheric ice content significantly hinders our ability to accurately model the climate. The current parameterizations used in GCMs vary by several multiplicative factors which has a significant impact on modeled hydrologic and radiative processes.

To address this problem, a retrieval algorithm was developed to determine cirrus cloud properties using a Bayesian integration method applied to submm and infrared data. As input to the algorithm, we will include cirrus cloud microphysics derived from in situ cloud probes, measured atmospheric profiles, and ISCCP (International Satellite Cloud Climatology Project) cloud products. By modifying these parameters we can simulate retrievals in different regimes (for example, tropical vs. mid-latitude) to determine the algorithm's performance for the variable conditions found globally. To simulate space-based measurements, the ISCCP data is sampled according to the orbit of the International Space Station.

We will report on the sensitivity of the retrieval simulations.