Publications

Kirthiga, SM; Patel, NR (2018). Impact of updating land surface data on micrometeorological weather simulations from the WRF model. ATMOSFERA, 31(2), 165-183.

Abstract
Land surface processes play a critical role in governing the surface energy partitioning and the atmospheric circulation within a climate system. Improper representations of present land state, particularly spatially specific fields such as land cover, topographical and biophysical parameters contribute to the uncertainty in the model's weather simulations extending from local to regional scales. The present study investigates the impact of superior land surface datasets on the performance of the Weather Research and Forecasting (WRF) model in simulating micrometeorological/near-surface weather, particularly sensible variables such as temperature, relative humidity, solar radiation and wind speed. The hypothesis is that the updated land surface datasets would help in improving micrometeorological forecasts over the domain comprising of Punjab, Haryana and Uttarakhand states in India. A land use land cover (LULC) dataset derived from Advanced Wide Field Sensor (AWiFS); an elevation dataset from the Shuttle Radar Topography Mission (SRTM), and a Leaf Area Index (LAI) based on the Moderate Resolution Imaging Spectroradiometer (MODIS), are used in model initialization. Performance evaluation of the model's simulation is done for controlled (default) and modified land boundary conditions with in situ weather from a network of automatic weather stations (AWS) operated by the Indian Space Research Organization (ISRO). In the modified run, the model more closely captured the temporal evolution of surface level temperature, relative humidity, wind speed, surface pressure and solar radiation. Improvement in 24-hr forecast ranges from 15 to 30% for these near-surface weather variables. Further testing of the model's performance on its capability to forecast 8-day micrometeorological weather variables revealed that the modified run gave consistent results. The average RMSE values for minimum and maximum temperature, wind speed, relative humidity and precipitation are 2.5 and 3 degrees C, 2 m s(-1), 18% and 3.5 mm, respectively. The modification helped in increasing the lead-time of the model's forecast by reducing the propagation error. Thus, this study emphasizes the fact that improved representation of land surface parameters has a definite effect on weather simulations at local to regional scales. For a country like India, where the feedback mechanisms between land and atmosphere are more prominent due to inherent climatic characteristics, it is critical to concentrate and improve on the inputs that represent the initial land state.

DOI:
10.20937/ATM.2018.31.02.05

ISSN:
0187-6236