Development of a High-Resolution Precipitation Climatological Dataset from the Climatology-Calibrated Precipitation Analysis (CCPA)

To fulfill the need for forecast calibration and verification studies and anomaly forecast guidance, a high resolution precipitation climatological dataset will be developed using the dataset of the Climatology-Calibrated Precipitation Analysis (CCPA). CCPA is a precipitation analysis developed at NCEP/EMC by statistically calibrating the multi-sensor Stage IV precipitation analysis over the Contiguous United States (CONUS) domain and makes its climatology closer to that of the rain gauge based estimate of CPC Unified Global Daily Gauge Analysis. It takes advantage of not only the higher reliability of the CPC Unified Global Daily Gauge Analysis but also the higher temporal and spatial resolution of the Stage IV dataset based on multi-sensor observations. However, CCPA spans only from 2002 to 2012 so that its climatology is based on a very short period of 10 years. This poses a challenge for generating precipitation climatology due to the lack of a long period of record to use as an analyzing period. Meanwhile, another outstanding challenge is the non-Gaussian nature of precipitation. Due to both aspects, climatological parameters which are derived from conventional methods are usually subject to bias and may not capture well the real frequency distributions. For both reasons, we will apply the method of L-Moments developed by Hosking and Wallis (1997) to produce a statistically reliable high-resolution climatological dataset of precipitation analysis over CONUS. The main advantage of the L-Moments, in comparison with conventional moments, is that L-Moments take into account linear statistics, making the estimation procedure more robust in the presence of extreme values in data. Another attractive feature is that L-Moments allow characterizing a wider range of frequency distributions even from small samples and exhibit lower bias than conventional moments. In this work, the preliminary steps and early results in the generation of the CCPA climatological mean (1st moment), standard deviation (2nd moment) and high moments using L-moments will be presented. The planned products can be extensively used for several studies on 1) QPF/PQPF calibration; 2) Hydrological applications which include initiating regional/global hydrological ensemble forecast model; 3) Model forecast evaluation; and 4) Generation of extreme forecast index (EFI) and anomaly forecast. In particular, future efforts will be made to mature the NCEP/GEFS's ability to provide precipitation anomaly forecast products.