The effects of land use change on the numerical modeling of regional climate and watershed runoff in the Great Lakes region.

Between 1895 and 1920 average air temperatures in the Midwest and United States declined by an amount similar to the increase observed from 1920 to 2000. The period 1895 to 1920 was also marked by land use changes from native forest and grassland to cropland. The investigation of a link between land use changes and climatic variability is the focus of this study. A mesoscale atmospheric model (MM5) was coupled with a puddle-modified Biosphere-Atmosphere Transfer Scheme (BATS) to model the regional climate. Verification simulations showed the BATS-modified MM5 was superior to the original MM5 in the prediction of surface water runoff and precipitation magnitude and distribution. To assess the effect of land use change on the summer climate of the Great Lakes region, four-month numerical experiments were completed using two land use scenarios, pre-settlement (1850) and current (1995). The results showed a significant modeled surface air temperature decrease in the latter period due to a repartitioning of the upward surface energy flux from sensible heat to evaporation. The modeled precipitation change was complex and appeared to be influenced very little by local land use change. However, there appeared to be a consistent modification of synoptic systems throughout the pre-settlement domain, with cold fronts and associated closed lows moving faster and isolated cold fronts moving more slowly. Clear, coherent mesoscale circulations formed over regions of abrupt, altered land use as expected due to surface heating but appear to have little effect on local precipitation. To assess the effects of future land use and climate change on the Huron River watershed, the BATS was further modified to include urban land use. The output from a global climate model was used to drive this version of the BATS. The simulations showed that the ten-year surface runoff would increase 28.9 inches from a current land use and climate scenario to a future land use and climate scenario. Factor separation analysis showed that 58% of this runoff increase would occur because of changing climate induced by higher intensity precipitation. The remainder of the increase was predominately due to forecasted land use changes.Ph.D.Physics, Atmospheric SciencePure SciencesUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/124735/2/3016797.pd