Sunday, 22 January 2012
The Sensitivity of a 6 and 12-hour WRF-ARW Forecasts for the Presidential Mountain Range to Model Input
Hall E (New Orleans Convention Center )
One of the purposes of Plymouth State University (PSU) NASA Experimental Program to Stimulate Cooperative Research (EPSCoR) Icing Assessments in Cold and Alpine Environments project is to predict ice accretion events using the Presidential Mountain Range of New Hampshire as the study area. A customized high-resolution Advance Research WRF (WRF-ARW) will be used to model icing events in an effort to establish the small-scale meteorological phenomena that are responsible for these events. To generate a successful forecast, the WRF-ARW relies heavily on the accuracy of model input that is used as the initial and boundary conditions. Errors that originate from the initial analysis and/or the lateral boundary conditions will eventually propagate through the model as time progresses forward. Thus, it was important to carefully choose the model input used as the initial and lateral boundary conditions for the PSU WRF-ARW.
The two models considered as the initial and lateral boundary conditions for the PSU WRF-ARW were the North American Mesoscale model (NAM) and the Global Forecast (GFS). To assess the quality the 6- and 12-hour forecasts generated from the PSU WRF-ARW, both sets of forecasts were compared to the true conditions of the atmosphere. The forecast accuracy of temperature, wind speed, relative humidity and heights for each model run were calculated in order to determine the quality of the outputted forecasts. Previous research has shown that these particular variables are essential for making an icing forecast. The accuracy of each model run was then compared to each other to assess the impact that different model input had on the WRF-ARW simulations. This paper will primarily report on this study, but will also briefly touch upon the future directions of using WRF-ARW for follow-on icing research.