Intercalibration of passive microwave rain products

Remote Sensing Systems processes passive microwave radiometer data from a variety of satellites: SSM/I on six DMSP satellites (F08, F10, F11, F13, F14, and F15), TMI on TRMM, and AMSR on Aqua and Midori-II. A great deal of effort has been spent accurately calibrating each sensor and intercalibrating all of the sensors. A standard algorithm has been developed that provides a suite of geophysical parameters, and the same algorithm is applied to the data from each sensor. Retrievals of sea surface temperature, surface wind speed, and water vapor made by the various sensors are in excellent agreement, and the retrievals have been extensively validated. Despite these facts, as we began to analyze rain rate retrievals, we found significant systematic discrepancies. The largest source of discrepancy, by far, was related to the resolution of the sensor. The resolution of the SSM/I measurements is nominally 32 km, while the other sensors are roughly 12 km. This difference in resolution has a profound effect on the retrieved rain rates through “the beamfilling effect”. The beamfilling effect does not impact the other parameters because the measured parameters are not nearly as spatially inhomogeneous as rain, and the relationships between those parameters and the radiometer measurements is not nearly as nonlinear as it is for rain. We will discuss the algorithmic changes we made in order to account for the beamfilling effect. We will present our rain rates both before and after correction, compare our rain rates against other climatologies, and present available validation. Finally, we will show the impact of the diurnal cycle on these measurements, and demonstrate applications for this now 18 year global record of rain rate over the oceans.