Abstract
A technique is described for using rain rates inferred from microwave low‐orbit satellite observations to objectively adjust rain rates inferred from geosynchronous‐orbit infrared (geo‐IR) satellite observations to produce monthly total rain maps for the region 40°N to 40°S. The adjustment is based on the spatially variable ratio of rainrate estimates from coincident microwave and infrared data, which is then applied to the full geo‐IR data set. The resulting “microwave‐adjusted geo‐IR rain estimate” has the (usually low) bias of the microwave estimates, together with the smoothness and temporal coverage of the geo‐IR data. Here, the microwave estimates from the Goddard Scattering Algorithm (GSCAT) are used to adjust the geo‐IR estimates from the Geosynchronous Precipitation Index (GPI) to form the Adjusted GPI (AGPI) technique. Verification against rain gauge analyses over water and land and subjective examination of the resulting maps and zonally‐averaged fields show that the AGPI estimates are superior to either the GSCAT or the GPI estimates by themselves for the four‐month PIP‐1 period of August‐November, 1987.
The GSCAT, GPI, and AGPI techniques are applied to the period of July 1987 to June 1988 to examine one annual cycle of tropical precipitation. Major climatological features are identified and compared with the three techniques. The AGPI has the spatial smoothness which is related to the frequent geosynchronous time sampling, but without the known biases of the GPI. The AGPI monthly rain total (integrated over 40°N to 40°S) has an annual cycle featuring a maximum in July and a minimum in January. The AGPI zonally averaged total rain for the year has its primary peak at 5–7.5°N and a secondary peak at 5–7.5°S with the annual integrated total being 85 mm mo−1.
