Comparing diabatic heating: conventional versus superparameterized climate model - related to the stochastic parameterization problem

The forecast error due to convective and radiative parameterizations remains a major uncertainty in weather and climate model predictions. A general approach to correct the model errors due to the parameterized tendency terms is to add informative stochastic perturbations to mimic the missing subgrid-scale physics. The first-step to this approach is to derive an informative/realistic perturbation time-series by calculating the differences between the model tendency and the realistic tendency. We can eventually obtain the probabilistic and statistical properties of the model errors through these perturbation time-series. The model tendencies are generated from the Community Atmosphere Model (CAM), and the realistic tendencies are from the superparameterized CAM (SPCAM), a 2-D cloud resolving model embedded CAM. Using the same CAM model allows one to easily compare different perturbation time-series throughout the spatial grid. In our previous study, we developed an efficient and reliable scheme by applying the stochastic perturbations in an idealized model framework. Our goal is to further employ our scheme with the help of the perturbation time-series to enhance the long term operational forecast for CAM.