WRF Model Evaluation for the Urban Heat Island Assessment under Varying Land Use/Land Cover and Reference Site Conditions

Abstract:

Urban heat island effect has been assessed using Weather Research and Forecasting model (WRF v3.5) coupled with urban canopy model (UCM) focusing on air temperature and surface skin temperature in the sub-tropical urban Indian megacity of Delhi. The estimated heat island intensities for different land use/land cover (LULC) have been compared with those derived from in-situ and satellite observations. There is a significant improvement in model performance with inclusion of UCM both for meteorological parameters (T and RH) and the UHIs. Overall, RMSEs for near surface temperature improved from 1.63°C to 1.13°C for urban areas and from 2.89°C to 2.75°C for non-urban areas with inclusion of urban canopy model in WRF. Similarly, index of agreement and RMSEs for mean urban heat island intensities (UHI) improved from 0.77 to 0.88 and 1.91°C to 1.60°C respectively with WRF-UCM. Hit rate from the model simulated mean heat island intensities using WRF model are 72 % for urban areas and 58 % for non-urban areas such as green areas and riverside areas. The corresponding values improved in WRF-UCM with a hit rate of 75% for urban areas and 72 % for non-urban areas. In general, model is able to capture the magnitude of UHI well though it performs better during night than during the daytime. High UHI zones and top 3 hotspots are captured well by the model.

The relevance of selecting a rural reference point for UHI estimation near the urban area is examined in the context of rapidly growing cities where nearby rural areas are transforming fast into built-up areas themselves and reference site may not be appropriate for future years. Both WRF and WRF-UCM simulated UHI shows satisfactory performance against benchmarks for the statistical measures with classical methodology using rural site as a reference point. Using an alternate methodology of considering a green area within the city having minimum temperature as a reference site worked satisfactorily only with WRF- UCM. In general, WRF-UCM always shows a better performance than WRF. Hence, inclusion of appropriate representation of urban canopies and landuse-landcover are important for improving predictive capabilities of the mesoscale models.