Challenges in forecasting the cloud-capped marine boundary layer and utilization of satellite data to improve mesoscale modeling

The objectives of the present study were to evaluate high-resolution mesoscale models in predicting the cloudy marine atmosphere and to develop new methods of assimilating the satellite data into the models to improve prediction accuracy. The study results provide significant guidance to optimizing and improving the real-time forecasts over the ocean, and in coastal zones in particular.; First, the study revealed that many of the model inaccuracies are due to coarse and imprecise initial and boundary conditions over the ocean obtained from the large-scale models. As a result, the success of the mesoscale model predictions of most of the observed cloudy marine layer fields is reduced by 20-50%, especially with regard to the moisture variables. Secondly, some of the operational turbulence parameterizations failed to couple the turbulence and cloud-radiative processes.; As pre-processors for four-dimensional data assimilation, we developed methods using adiabatic constraints to obtain the heights of the low-level cloud products from the satellite imagery, and found that the accuracy of the cloud-top height estimates is better than in other similar methods. The cloud-top heights combined with climatological evidence and surface observations helped us to generate more realistic three-dimensional temperature and moisture data in the coastal and offshore regions. Assimilation of the boundary layer proxy data into the initialization of the Mesoscale Model 5 (MM5) and Weather Research and Forecasting (WRF) models significantly increased the prediction accuracy of the marine layer depths by 30-35%, inversion strengths by 80%, and turbulence kinetic energy by 40%.