Improving Mesoscale Weather Simulations Through Updated Land Use and Vegetation Information

Changes in land use and land cover(LULC) directly affect physical characteristics of the planetary boundary layer (PBL) - such as temperature, humidity, and wind – by varying the energy and moisture balance between the surface and the atmosphere. As a result, improved characterizations of LULC within numerical weather prediction (NWP) models should lead to increased accuracy of simulations. This is especially true during the growing season over agricultural areas, when increased vegetation coverage and leaf area index (LAI) enhance evapotranspiration (ET) and associated processes such as convective rainfall. Many NWP models utilize a long-term average of vegetation data as a static input for surface boundary conditions, which may be acceptable for simulations at the synoptic scale; however, for mesoscale simulations this can lead to inherent errors in the accuracy of the resulting simulations. In the same way that assimilation of atmospheric data is critical for improved accuracy of NWP model output, near-real-time characterization of surface conditions is necessary to enhance forecasts of surface-based weather processes and PBL characteristics. Building on this idea, the objective of this study is to quantify the influence of updated LULC and LAI data on mesoscale NWP model simulations over an agricultural area through the course of an annual growing season. The approach involves modification of the geostatic files within the Weather Research and Forecasting (WRF) model using LULC and LAI products from the Moderate Resolution Imaging Spectroradiometer (MODIS) environmental satellite platform at seasonal and weekly time scales, respectively. In comparison with the current WRF vegetation data, the MODIS near real-time LAI products add realistic vegetation characteristics and phenology to the model simulations. The simulation approach involves four independent model runs: (1) standard WRF LULC and LAI, (2) updated LULC from MODIS with standard WRF LAI, (3) standard WRF LULC with updated LAI from MODIS, and (4) updated LULC and LAI from MODIS. All simulations will be performed over the 2018 growing season (April 15 – October 1), with data from May 1 – October 1 being used for analysis (15-day model spin-up) over a 5-km nested domain centered over the Lower Mississippi River alluvial valley (LMRAV) with a 1-km domain centered over the southern LMRAV in Mississippi and Arkansas. Based on this model framework, the seasonally averaged MODIS LULC data from 2017 (MODIS last available land use data) will be used to replace the WRF data, while the LAI data will be updated every eight days using data from 2018. It should be noted that the MODIS LULC data will be used to update all LULC fractions within WRF, not just the dominant land use type. As the primary aim of the project is to investigate the improvement of WRF simulations of lower atmospheric properties and surface-based processes, analysis will focus on the accuracy of 2-m temperature and humidity, 10-m wind speed, convective precipitation distribution and intensity, and soil moisture and temperature. Gridded precipitation data from the multi-sensor Stage IV product as well as point data from the Soil Climate Analysis Network (SCAN) and NWS ASOS stations will be used for verification of soil and atmospheric properties, respectively. The accuracy of precipitation simulations will be the primary focus as this information is vital for decision makers in agricultural fields in terms of field management, crop planning, and irrigation scheduling. Results of this study will not only show how updated LULC and LAI information from environmental satellites can improve mesoscale NWP model performance, but also how important surface-based processes are on lower-atmospheric characteristics over larger agricultural regions. By improving both model accuracy and the general scientific understanding of the role of land use and vegetation on atmospheric properties, future predictions can be improved to benefit agricultural productivity and water resources management.