Verification of cloud production in the Community Atmosphere Model: A comparison of two data assimilation techniques

Abstract

Clouds play an important role in regulating our climate, and it is vital that we are able to accurately simulate them in global climate models. Data assimilation can be used to force the model towards simulating a specific historical event, and then the clouds simulated can be compared to those observed during the event. Two data assimilation methods, Newtonian relaxation and Kalman filtering, are used to force the Community Atmosphere Model version 6 towards the state of the atmosphere that was observed in July of 2013. The use of the two methods produced significantly different results. The results produced using Newtonian relaxation indicate the Community Atmosphere Model simulates clouds reasonable well. The results produced using Kalman filtering support the conclusion that there may have been an error in the way the model was set up for that run.