Lake Process Simulations with a Regional Climate Model

Lake-atmosphere interactions are vital to weather and climate predictions over lakes and nearby regions. To dynamically simulate lake processes and lake-atmosphere interactions, a one-dimensional, physically based lake model was coupled with the next generation Weather Research and Forecasting (WRF) model to improve predictions of lake-effect precipitation over lakes and surrounding areas. Initial results from the coupled WRF-Lake model showed a greatly exaggerated seasonal cycle of lake surface temperature (LST), especially in deep lakes (e.g., Lake Superior, with a depth greater than 50 m). Realistic LST simulations were achieved by adjusting the eddy diffusivity and surface roughness length for lakes in the model. In addition, multi-year simulations were performed with the calibrated WRF-Lake model at 10 km resolution for the Great Lakes region over the period of 2003-2008. The model initial and lateral conditions were provided by the 32 km resolution North American Regional Reanalysis data. The results reveal that the simulated LSTs agreed well with buoy observations and Moderate Resolution Imaging Spectroradiometer satellite data. These simulations also generated more accurate lake-effect precipitation than those produced by the recent version of WRF, without including a lake scheme.