Requirements for a robust precipitation constellation

Over the last 15 years the constellation of satellites carrying passive microwave (PMW) sensors has grown to a mature collection of almost a dozen satellites at any given time. Increasingly, a broad range of science and user communities have come to depend on the quasi-global precipitation analyses that intercalibrate and merge these individual PMW precipitation data streams. At present, the constellation of precipitation-relevant conical and cross-track scanning multichannel PMW instruments depends on many satellites past their design life and in continued operation by the responsible agencies. The Group on Earth Observations (GEO) Water Strategy and subsequent discussions in Coordinating Group for Meteorological Satellites (CGMS) and Committee on Earth Observing Systems (CEOS) have raised the issue of how a robust future precipitation constellation should be constructed. We discuss how retrievals are impacted by sensor resolution and channel diversity, the observation interval, and the use of a quasi-operational satellite precipitation radar for calibration. Specifically: 1) Sensor footprints larger than about 5-10 km start to introduce significant amounts of non-linearity in the retrievals, the so-called beam-filling problem. 2) Channel diversity has been shown to be necessary for covering the range of precipitation rates and types (liquid vs. solid). As well, diverse polarization at a given frequency is also important. 3) An observation interval less than three hours for every time around the day barely accommodates the required revist times for cloud-scale precipitating systems. 4) The precipitation radars on the precessing TRMM and GPM Core Observatory satellites have demonstrated the utility of routine calibration for precipitation estimates across all the PMW sensors (and in the case of GPM this is done for most of the Earth's climate zones). Such considerations are critical to the discussion on how to shift to a new, more diverse generation of precipitation-relevant sensors while preserving the characteristics that provide (and support continued innovation of) quality PMW retrievals and valueadded products that many users find attractive.