Extended Abstract (2.4M)1.10
Using new remotely-sensed datasets in MOS development: A step toward true high-resolution statistical quantitative precipitation forecasts
Mark S. Antolik, NOAA/NWS, Silver Spring, MD
The availability of remotely-sensed meteorological data from a variety of sources now makes it possible to contemplate the development of true high-resolution statistical forecast systems. Such statistical systems will provide guidance tools to help the forecaster interpret high-resolution numerical weather prediction model output and provide detailed precipitation information to support the new National Weather Service (NWS) National Digital Forecast Database. Accordingly, the NWS Meteorological Development Laboratory (MDL) is now developing a new high-resolution gridded Quantitative Precipitation Forecast (QPF) system by applying the Model Output Statistics (MOS) technique to a nationwide mosaic of Quantitative Precipitation Estimates (QPE) collected at the National Precipitation Verification Unit (NPVU).
The NPVU QPE dataset is comprised of multisensor precipitation estimates produced routinely by the Hydrometeorological Analysis and Support units of the NWS River Forecast Centers (RFCs) for ingest into operational streamflow models. These precipitation estimates are collected by NPVU for 6-h intervals on the 4-km Hydrologic Rainfall Analysis Project (HRAP) grid commonly used in NWS operations. Over the eastern United States, these estimates are based primarily on WSR-88D radar data debiased by using available gauge observations. In the mountainous West, however, these estimates are produced by using gauge observations interpolated to the HRAP grid with the help of high-resolution climatology and the Mountain Mapper software currently in operational use at the three western RFCs.
This paper describes initial steps in the design and development of a new MOS QPF system based on the NPVU QPE dataset. Discussion focuses on the technical challenges associated with producing nationwide, gridded MOS QPF guidance at high resolution. Among these issues are the nonhomogeneous nature of the proposed predictand dataset, the evolving characteristics of operational QPE processing techniques and data quality control efforts at the RFCs, and the sheer volume of the computations required to produce this guidance over the entire CONUS. We also discuss how this system fits into the context of broader efforts to produce centralized MOS forecast guidance on high-resolution grids as well as some implications for hydrologic operations.
Session 1, Forecast Systems (Room 602/603)
Monday, 12 January 2004, 9:00 AM-4:30 PM, Room 602/603
Previous paper Next paper