6th International Conference Energy & Meteorology: Abstract Submission

Shortwave and longwave radiation forecast uncertainty of the HRRR and HRRRNEST NWP Models during specific meteorological regimes across four sites in the Columbia River Basin during the Wind Forecasting Improvement Project (WFIP-2)  (822)

Kathleen Lantz 1 2 , Chuck Long 1 2 , Gary Hodges 1 2 , Emiel Hall 1 2 , James Wendell 2 , Joseph Olson 1 2 , Jaymes Kenyon 1 2 , Dave Turner 2
  1. CIRES, University of Colorado, Boulder, CO, US
  2. NOAA, Boulder, CO, United States

Objective and Overview:  The second Wind Forecast Improvement Project occurred for 18 months from Oct, 2015 – Mar, 2017 in the Columbia River Basin area in the Northwest U.S.  The primary goal was to improve wind forecasts in complex terrain.   This effort involved multiple U.S. agencies, industry, and Universities with a large suite of instrumentation including vertical profiling wind radars (WPR), sodars, lidars, and additional measurements including shortwave and longwave radiation, sensible and latent heat for understanding the development of low-level winds. 

Methodology:  In this study we evaluate the diurnal and seasonal variability of in-coming and out-going surface shortwave (SW) and longwave (LW) radiation, and cloud cover from the HRRR (3-km) AND HRRRNEST (750-m) forecasts using surface observations.  This study looks at the variability in radiation quantities across four sites within the study region of the Columbia River Basin and investigates how well the magnitude and variability of these observations is captured by the HRRR and HRRRNEST NWP models using standard statistics (RMSE, MBE, MAE).  The HRRR was targeted for specific improvements in model physics such as scale-aware aspects of turbulence parameterizations (PBL + shallow cumulus scheme), representation of sub-grid scale clouds, and land-surface physics.  This study looks at how the improvements in model physics improved the standard statistics across four 6-week seasonal reforecast runs and two 2-week retrospective runs.  This study also looks at the how the forecast uncertainties vary under specific meteorological regimes, e.g. persistent cold pools and mountain gap events. 

Discussion:  There were significant improvements in the radiation quantities from the control runs to the experimental runs with the improved model physics.  At the Wasco, OR site, the MBE in downwelling shortwave radiation in the reforecast runs improved by 10-73% across the 4 seasons, and downwelling longwave radiation by 30-78%.  Statistical results from across the sites and future plans will be presented.