Precipitation Forecasts Using The WRF-ARW and Wrf-nmm Models During The Hmt-west 2006 and 2007 Winter Experiments

Abstract

The NOAA Hydrometeorological Testbed (HMT) program (http://hmt.noaa.gov) is a long-term project, which links the research community with the operational weather and hydrological forecasting. The HMT-West winter experiments focus on the winter storms over the American River Basin (ARB) of Northern California. The NOAA/ESRL/GSD operated a time-lagged multimodel ensemble system for providing weather forecasting during the HMT-West 2006 and 2007 campaign. Based on the HMT-West 2006 operations, ensemble forecasts were reconfigured for the HMT-West 2007 real-time forecasts. Multiple microphysical schemes and dynamic cores were used in weather research and forecasting (WRF, http://www.mmm.ucar.edu/wrf/users/) model. The four model runs include the WRF with Nonhydrostatic Mesoscale Model (NMM) dynamic core, which uses the Ferrier microphysics, and with the advanced research WRF (ARW) dynamic core, which uses the Ferrier, Thompson, and Schultz microphysics. All models were initialized per 6 h with the Local Analysis and Prediction System (LAPS) diabatic hot start and run to 30 h at 3-km resolution over Northern California (150 x 150 grids). Convective schemes were not used at 3–km resolution. Boundary conditions were from 40-km North American Mesoscale (NAM) model. The 6-h/24-h quantitative precipitation forecasts (QPF) and probabilistic QPF (PQPF) from individual models and time-lagged multimodel ensembles were analyzed for the selected intensive operation periods (IOPs) during the two winters. The NCEP Stage IV 6-h precipitation analyses were selected as the observations. Both microphysical schemes and dynamic cores showed high impacts on precipitation forecasts over this mountainous area. Time-lagged multimodel ensembles can effectively use the previous runs, and the LAPS system avoids the “spin-up” problem. This ensemble system improves the QPF compared to individual models, especially for 6-h accumulations.