Evaluating Satellite-Based Cloud Persistence and Displacement Nowcasting Techniques over Complex Terrain

Abstract

he Moderate Resolution Imaging Spectroradiometer (MODIS) sensor aboard the Terra and Aqua satellites allows roughly for the same region of Earth to be sampled twice in a nowcasting time frame. Using the MODIS cloud mask at 5-km resolution and radiosonde data, cloud nowcasting methods were evaluated from December 2008 through November 2009 over Utah and southwestern Wyoming (207 cases), a region with a variety of terrain. The hypothesis that a combination of persistence and displacement techniques (i.e., a hybrid technique) in regions of complex terrain should yield a better forecast than either method alone is tested; clouds forced by complex terrain often appear to persist along topographic barriers. A new forecast skill evaluation method is introduced, designed to equate correct cloudy and correct clear areas. The persistence nowcasting method demonstrated the highest skill [as much as 10% critical success index improvement (CSI) over other methods]. However, all nowcasting techniques performed similarly during the summer months (~65% CSI). It is concluded that cloud nowcasts at 5-km resolution over complex terrain in the forecast area, using no model wind or moisture data, do not improve upon a persistence nowcast. However, because these basic nowcasting methods are computationally fast, educated decisions on minute to hour time scales can be made nearly instantaneously.