Investigating the relationship between urban landuse and precipitating convective systems over the Atlanta region

This study uses Weather Research Forecasting model (WRF version 2.2) to simulate convective precipitation characteristics over the Atlanta region. The study identifies and quantifies the role of urbanization in the initiation of precipitating convection over the region.

Based on analysis of radar composites and surface maps, rainfall activity over the region throughout the 17th August 2002 was selected as an event with likely urban forcing. With the city of Atlanta at the center, three domains of horizontal grid resolution sizes of 30km, 10 km and 3.3 were nested as domains 1, 2 and 3 respectively, each with 31 vertical levels.

The basic atmospheric physics packages used were Kain-Fritsch (new Eta) scheme for cumulus cloud parameterization in domains 1 and 2, Ferrier microphysics in domain 3, while the Noah Land Surface Model (LSM) represented the land surface characteristics based on the United States Geological Survey (USGS) land use dataset. The model is initialized with gridded analyses data from National Center for Environmental Prediction (NCEP -Eta/NAM), 40 km resolution-every 6hrs.

The 9hr control (urban) simulation beginning 3hrs prior to the observed storm initiation shows that the model does reproduce the convective evolution of the case day. The alternative (nourban) case indicates that removing the city of Atlanta (in domain 3) causes distinct differences in accumulated rainfall totals and influences the temporal and spatial evolution of the explicitly resolved precipitation.

Preliminary results indicate varying sensitivity of the urban-enhanced convection to specific model parameters such as roughness length, surface albedo, soil thermal conductivity, volumetric heat capacity and greenness factor.