A simple method to retrieve 3-hourly estimates of global tropical and subtropical precipitation from international satellite cloud climatology program (ISCCP) D1 data

Algorithms to estimate rainfall from passive microwave or optical data from polar-orbiting satellites are limited by poor temporal sampling and are be st suited to produce estimates integrated over periods of one month or more . There are numerous applications in the atmospheric sciences in which rain fall estimates are required at a much greater frequency. These can be deriv ed from geostationary satellite infrared data, but currently no global arch ive of such products exists. This paper presents a simple technique to reco nstruct Geostationary Operational Environmental Satellite Precipitation Ind ex (GPI) estimates of rainfall over the global Tropics and subtropics at 3- hourly, 2.5 degrees resolution from cloud-top temperature statistics contai ned in the extensive International Satellite Cloud Climatology Project D1 d ataset. It is shown that the Reconstructed GPI (RGPI) estimates correlate v ery strongly with the GPI and have minimal bias, irrespective of the integr ation period selected or the underlying surface type. Comparison with the i ndependent NASA WetNet PIP-3 surface rainfall validation data shows that th e RGPI estimates of rainfall composited over monthly periods march the vali dation data with accuracy very similar to that of the GPI and are comparabl e to many passive microwave algorithms. Both the RGPI and GPI estimates of rainfall match the validation data more closely over the tropical Pacific O cean than over the tropical and subtropical land masses where a positive bi as is apparent. With 3-hourly temporal resolution, the RGPI represents a us eful new resource for climate studies.