The Impact of Low Clouds on Surface Shortwave Radiation in the ECMWF Model

Abstract The long-term measurement records from the Atmospheric Radiation Measurement site on the Southern Great Plains show evidence of a bias in the ECMWF model’s surface irradiance. Based on previous studies, which have suggested that summertime shallow clouds may contribute to the bias, an evaluation of 146 days with observed nonprecipitating fair-weather cumulus clouds is performed. In-cloud liquid water path and effective radius are both overestimated in the model with liquid water path dominating to produce clouds that are too reflective. These are compensated by occasional cloud-free days in the model such that the fair-weather cumulus regime overall does not contribute significantly to the multiyear daytime mean surface irradiance bias of 23 W m−2. To further explore the origin of the bias, observed and modeled cloud fraction profiles over 6 years are classified and sorted based on the surface irradiance bias associated with each sample pair. Overcast low cloud conditions during the spring and fall seasons are identified as a major contributor. For samples with low cloud present in both observations and model, opposing surface irradiance biases are found for overcast and broken cloud cover conditions. A reduction of cloud liquid to a third for broken low clouds and an increase by a factor of 1.5 in overcast situations improves agreement with the observed liquid water path distribution. This approach of combining the model shortwave bias with a cloud classification helps to identify compensating errors in the model, providing guidance for a targeted improvement of cloud parameterizations.