On the impact of urbanization on summertime thunderstorms in Atlanta: Two numerical model case studies

There is increasing observational evidence that urban land cover can have a significant effect on precipitation variability. Atlanta, because of its rapid urbanization, has been a focus for several observational studies on urbanization and rainfall. Yet. there is a lack of numerical model studies in the literature to examine physical processes linked to the Atlanta "urban rainfall effect" or URE. This study employs the Weather Research and Forecast (WRF) model to simulate convective precipitation for two cases: 17 August 2002 ("an urban interaction case") and 26th July 1996 (an urban "initiation" case). Case I was chosen based on analysis of radar composites and surface maps which identified it as an event likely to have some urban forcing with minimal large scale forcing; while Case 2 was based on a previous observational study indicating that the storm was initiated by urban heat island induced convergence zone. The control (URBAN) simulations begin 3 h prior to the observed storm initiations and reveal that the model captures the convective evolution of the cases. The alternative (NOURBAN) simulations indicate that removing the city of Atlanta causes distinct differences in the temporal and spatial evolution of the explicitly resolved precipitation. However these differences point more to the likelihood of modification rather than initiation of the convective systems. Time series of cumulative rainfall totals indicate that the explicitly resolved rainfall is initiated at the same time in both the URBAN and NOURBAN simulations. The rainfall initiation time even within sub-sections of the domain is the same for the URBAN and NOURBAN scenario. Rainfall amounts downwind of the city are higher by 10% to 13% within a strip 20-50 km east of the city, for the URBAN schemes in comparison to NOURBAN. (C) 2008 Elsevier B.V. All rights reserved.