Japan Meteorological Agency Adopts DOE Funded Cloud Science to Improve Global Precipitation Forecasts

Summary

Scientists in DOE s Atmospheric Radiation Measurement (ARM) program developed a new way to determine when and by how much deep clouds should form in a mathematical model. This new method has recently been incorporated in the global weather prediction model of the Japanese Meteorological Agency (JMA) to improve its precipitation forecast. The new cloud formulation relates the occurrence and magnitude of deep clouds to the rate at which the large-scale atmospheric circulation creates instability in the atmosphere while the old formulation was solely dependent on cloud instability alone. In many climate models, deep cloud formation happens much frequently than observed in nature. Also, the convection intensity of these deep clouds in those models is lower than that of clouds in nature. The salient features of the new formulation are that it helps to(1) reduce unrealistic frequent formation of deep clouds and (2) improve the intensity of convection, when convection occurs in the model. Both of these changes reduced biases that are common in many climate models. In the past, this formulation has already improved climate simulations in the NCAR climate atmospheric model (CAM).