Traditional Statistical Measures Comparing Weather Research and Forecast Model Output to Observations Centered Over Utah

摘要

The Model Assessment Project automated legacy scripts produce model validation statistics using the National Center for Atmospheric Research (NCAR) Model Evaluation Tools (MET) software. The project also accessed the Meteorological Assimilation Data Ingest System (MADIS) mesonet observation data to augment the legacy National Centers for Environmental Prediction (NCEP) PrepBUFR METAR observational data to provide a twenty-fold increase in the number of observations to compare with model forecasts. MET Point-Stat was used to generate statistics based on the point differences between the model forecast and the observations. The statistics were then aggregated to produce a summary of results for twenty case study days for nine surface meteorological variables. The model forecasts were generated by the Nowcast Modeling Project in support of a request from Air Force Weather Agency (AFWA) to compare the performance of the Advanced Research Version of the Weather Research Forecasting model (WRF-Advanced Research WRF ARW) using seven different model parameter settings. These model runs were executed over the Dugway, Utah area and evaluation of the statistical output is discussed with some graphical examples.