Increasing the accuracy of the WRF-ARW numerical weather prediction model using Corine Land Cover and JAXA data

Abstract

The geographic data play an important role in atmospheric and weather forecasting. They influence input physical parameters defining the energy balance and heat and moisture fluxes in the planetary boundary layer. The goal of this paper is to increase the accuracy of the widely used Weather Research and Forecasting (WRF) numerical prediction model using a more detailed, 100 m resolution Corine Land Cover 2018 (CLC) and Japanese Space Agency (JAXA) elevation data in comparison to global data from the United States Geological Survey (USGS) containing the MODIS LULC land cover and elevation data. The WRF-ARW variant of the model is applied to a study area of 50x50 km in the Košice region at a microscale level with a spatial resolution of 250 m for June 24, 2020. The GIS4WRF module in QGIS and other GIS tools are used to perform modelling and input and output data preparation and result analysis. The results for both datasets are compared to data measured in local weather stations and statistically evaluated. The use of higher-resolution geographical data improves the overall accuracy of the model for this microscale level.