Evaluation of FV3 Model Using Satellite Brightness Temperature Data

Incorporation of physics schemes from a repository, such as the Development Testbed Center's Common Community Physics Package (CCPP), into a new forecast model requires testing and evaluation of the connections between the schemes. This study uses GOES-16 satellite data from the National Oceanic and Atmospheric Administration (NOAA) to assess the forecast skill of the updated Next Generation Global Prediction System, with particular regard for the cloud related atmospheric physics parameterization schemes. Evaluation metrics were produced using the Model Evaluation Tools (MET) package developed by the National Center for Atmospheric Research. With respect to season and cloud altitude, the analysis of the output statistics revealed that the upgraded forecast model produced the least accurate and inconsistent results during summer, rather than winter, for higher, colder clouds. Concerning the coldest clouds, the model tended to overpredict cloud presence in the winter and underpredict it in the summer. Future diagnostic work is needed to identify the cause of these problems.