This numerical study examines the impact of urban growth and release of aerosols, moisture, and heat on precipitation for Fairbanks, Alaska, a remote city at high latitude. The remote location allows atmospheric changes to be attributed to the city and permits examination of their regional-scale impact. The fifth-generation Pennsylvania State University-NCAR Mesoscale Model (MM5) with three nested, interacting domains was run over Alaska. An analysis of variance separates the variance in precipitation ascribable to one urban causal factor from variance ascribable to other urban factors. Statistical analysis using an unreplicated 2(4)-factorial design detected significant precipitation modification by individual urban effects, plus interactions among effects. Release of aerosols, release of moisture, and concurrent urban growth and aerosol release statistically significantly affect downwind precipitation with 95% confidence. The significant response to urban aerosol release results from directly caused shifts in cloud microphysical processes. For the other simulations, altered buoyancy indirectly initiates shifts in cloud microphysical paths. Additional moisture from urban sources contributes to increased downwind precipitation. Locations of maximum increase in precipitation vary for the various urban effects because of different mechanisms involved. All urban effects investigated slightly influenced accumulated precipitation predicted for the innermost domain, but not for the outer domains; prediction efficiency depends on grid increment size because processes are resolved on the finer grid that do not appear on the coarse grid. Changes in precipitation can have both positive and negative ecological and societal impacts; identifying which urban effects might cause precipitation changes can help in cost-effective planning to ameliorate negative consequences.
