Evaluation of OLR-based CPC high-resolution precipitation in GLDAS re-run experiment

Improvement in the Global Land Data Assimilation System (GLDAS) experiment is driven by the need to create more climatologically consistent data sets. Currently GLDAS includes upgraded versions of Land Surface Models (LSMs), enhanced global meteorological forcing data sets, more robust soil moisture initialization, updated model specific parameter sets and an advanced snow data assimilation scheme. In this experiment, we are using one of the latest high-resolution precipitation data sets, which has utilized a long-term climatology of outgoing longwave radiation (OLR) estimated from satellite observations to produce a blended precipitation product. This data set is 0.25-degree at a daily interval, which is further downscaled to GFS T1536 resolution through a hybrid-stochastic precipitation downscaling method. The OLR-based blending of precipitation is noticed to be capable of providing a globally uniform and temporally stable high quality product, similar to geostationary satellite IR-based estimates over the tropics and to estimates based on microwave scattering observations over extra-tropical areas. We will present comparative results of fluxes and water budget components from a newly run GLDAS experiment and previous GLDAS and GLDAS2 runs.