Exploring Coupling to Improve Lake-Effect Snow Forecasts

While much is known about lake-effect snowstorms, forecasts are still typically based on bulk parameter space and pattern recognition. Similar to their deep convective counterparts, predicting the existence of the phenomena is reasonably well-handled; however, specificity of each convective episode still poses significant prognostic challenges. Application of numerical weather prediction models at higher resolution has typically resulted in forecast skill improvements of mesoscale processes with the exception of lake-effect snow prediction. Much of the lag is due to insufficient resolution of forecast model applications. Moreover, construction of physics parameterizations - especially boundary layer, land surface, and microphysics schemes - are challenged by the extreme conditions presented by lake-effect convection. In addition to the challenges posed to numerical model construction and configuration, variations in surface boundary conditions - specifically lake-surface temperature and ice coverage - can dramatically alter the evolution of lake effect snow band structures. Furthermore, those surface boundary conditions also change through the course of an Arctic airmass episode, thus lending to a coupled modeling approach. However, coupled modeling is yet to be applied with operational systems to the Great Lakes system. In response, ongoing initial work between NOAA/NWS, NOAA/OAR/GLERL, and NOAA/OAR/ESRL/GSD seeks to demonstrate the potential value of coupling through iteratively informing the developmental version of the operational hydrodynamic model (Finite Volume Community Ocean Model, FVCOM) with an operational short-term mesoscale atmospheric model (High-Resolution Rapid Refresh, HRRR) to address improving lake-effect forecasts. Early results from the ongoing lake-effect improvement project will be presented including leveraging flux-tower measurements to isolate deficiencies in boundary-layer physics parameterizations in lake-effect applications.