A technique for understanding effects of urbanization on local precipitation using ground station data

It has long been hypothesized that urban areas affect local weather patterns. Urban areas typically have higher temperature than the surrounding areas (known as the urban heat island or UHI), higher roughness, and lower surface moisture than rural areas. These differences lead to zones of surface convergence and uplift which may increase the strength of passing storms (Shem and Shepherd, 2009, Atmospheric Research). Since these zones are expected to generate at the edge of an urban area, precipitation is expected to be higher near the edges of a city than in the city center or the surrounding rural region. As an urban area expands, the region of intensification is expected to move outward from the city core. Furthermore, larger cities are expected to have stronger effects than smaller ones (Ashley et al, 2011, Climatic Change). One approach to evaluating such effects is observation of urbanization and precipitation changes over time. While short-term analysis may be conducted using a combination of satellite and radar data, such datasets do not extend much prior 2000. Many urban areas underwent large changes in decades prior to this period. Station data are available for long periods of time, but suffer from lack of density and are prone to uncertainty due to station location change. Analysis in rain type change among stationary stations is statistically difficult due to the heavy influence of individual events over short time periods (Diem and Mote, 2005, J. Applied Meteo.). These problems are compounded by the fact stations do not include information about land use type, which may change over time, and is an important factor in urban effect analysis. For this study, an algorithm has been developed to identify land use type for the greater Washington, DC metropolitan region for the period of 1970-2010 using temperature data from stations in the NWS Cooperative Network. Station land type have been identified according to development level (urban core, suburban, rural etc.). The changes in land type for given regions is examined to determine urbanization over time. Meanwhile, differences in precipitation means for each land-use group are examined to discern regions of high rainfall relative to urbanization. Station land type for 2010 determined by algorithm are compared to UHI derived from surface skin temperatures measured by NASA's satellite-mounted Moderate Resolution Imaging Spectroradiometer (MODIS) instrument, and precipitation patterns are compared to radar reflectivity from the NEXRAD LWX WSR-88D Doppler radar at Sterling, VA. If storms are enhanced by urban-edge updrafts and these updrafts are linked to urban size, it is expected that the highest amount of rainfall will occur for stations with mixed urban/rural influence (on the urban edge) and the rainfall enhancement in these areas will increase as the city expands. Confirmation of these hypotheses will enhance the knowledge of urban effects on precipitation. Such knowledge of local precipitation is vital to proper agricultural and urban planning and knowledge of where precipitation is likely to intensify will aid forecasters in issuing severe weather warnings.