A Multisensor Analysis of Light Preciptation Regimes

ABSTRACT
It is anticipated that the temporal resolution afforded by the constellation-based Global Precipitation Measurement (GPM) mission will foster extensive use of satellite-based rainfall estimates for land-surface hydrologic applications such as water resource management, flood forecasting, drought monitoring, and agricultural applications. Unlike their oceanic counterparts, however, operational passive microwave rainfall algorithms over land tend to be largely empirical due to the tenuous connection between observed brightness temperature and rainfall. Algorithm "improvement" thus consists generally of adjustments in rainfall screening and regional tuning of brightness temperature-rainfall relations to improve the statistical agreement with surface observations on monthly timescales while fundamental changes to more physical schemes that make use of the wealth of ancillary satellite data remain untapped. This proposal aims to develop such a physical framework using TRMM/GPM field experiment data to construct more physically based procedures for use by the land algorithm. As an initial step, we aim to exploit TRMM/GPM field experiment data and cloud resolving model output to assess the information that is potentially available and how to optimize its use within operational algorithms. As such, the effort is intended to establish a parallel path to operational land-based precipitation algorithms that will offer physical insight that can be incorporated into these approaches in a manner similar to that through which their oceanic counterparts evolved over time. In this way the findings of the proposed research will seek to address the fundamental flaw in conventional land-precipitation algorithms namely that, even with meticulous tuning, bulk statistical relationships between ice scattering aloft and rainfall at the surface, cannot provide sufficient accuracy at the high spatial and temporal scales required for hydrologic applications seeking to model high impact events that, by definition, do not conform to longer-term statistics.