LAUSR

The Earth's atmosphere is highly coupled between the vertical layers and the surface. An understanding of circulations in the atmosphere is important for developing models and improving weather forecasting. The latent heat produced in the atmosphere is one of the key driving forces of these circulations. It is therefore very important to estimate the latent heat in the atmosphere accurately, especially in thunderstorm clouds, which have proved to be one of the major sources of gravity waves in tropical regions. The current space‐based latent heat retrievals are limited to precipitation‐based estimation which cannot define the complete structure of a thunderstorm where precipitation is not the main indicator of the severity. A novel method is proposed in this study which retrieves the latent heat profiles of thunderstorm clouds using CloudSat W‐band radar profiles. A realistic database of simulated thunderstorm events developed using the Regional Atmospheric Modelling System – Cloud Resolving Model (RAMS‐CRM) is compared with observations using the Bayesian Monte Carlo method to derive an estimate with an uncertainty analysis for each estimate. The method is validated in the southeast Asian region with European Centre for Medium‐Range Weather Forecasts driven RAMS‐CRM profiles. The algorithm performance on observation data using linear fit, regression and bias analysis is discussed. A case study retrieval is also performed to demonstrate the retrieval on real CloudSat data.