Addressing Uncertainties in Cloud Top Height Verification with Synthetic Brightness Temperatures

Cloud top verification is inherently difficult due to large uncertainties in the estimates of observed cloud top height.  In this work, synthetic brightness temperature forecasts from the Coupled Ocean Atmosphere Mesoscale Prediction System (COAMPS^®#) are compared to forecast cloud top heights to investigate potential errors and develop filters to remove optically thin ice clouds.  Results from a statistical analysis reveal that up to 50% of the clouds with brightness temperatures as high as 280 K are actually optically thin cirrus.  The filters successfully removed most of the thin ice clouds, allowing for the diagnosis of very specific errors.  The results indicate a strong negative bias in mid-tropospheric cloud cover in the model, as well as a lack of land-based convective cumuliform clouds.  The model also predicted an area of persistent stratus over the north Atlantic which was not apparent in the observations.  In contrast, high cloud tops associated with deep convection were well simulated as were tradewind cumulus in the Sargasso Sea.

^#COAMPS^® is a registered trademark of the Naval Research Laboratory.