Evaluation of Rainfall Characteristics in the Goddard Multi-scale Modeling Framework (GMMF)

It is well known that precipitating and convective cloud systems play important roles for climate, but they are difficult to deal with in traditional global models with relatively coarse resolution ( 10 to 100 km). The cloud-permitting multi-scale modeling frameworks (MMFs) that aim to better represent these cloud systems must demonstrate their capability in accurately simulate the spectrum of precipitating system sizes and rainfall characteristics. To globally quantify the populations, horizontal and vertical structure, and characteristics of simulated cloud and rainfall systems in the Goddard MMF (GMMF), statistics of precipitation features (PFs) are derived from model radar reflectivity fields using the Goddard Satellite Data Simulator Unit. The simulated rainfall characteristics and horizontal and vertical structure of PFs are evaluated against TRMM satellite products. The results shown similar lognormal size distribution of PFs, rainfall characteristics, geographical distribution and variability as satellite observations. Furthermore, a series of model simulations were performed to systematically examine the impacts of the domain size and model grid resolution of the embedded CRMs on simulated PFs. The results indicate that embedded CRMs with large domain and fine resolution tend to produce better simulations compared to those simulations with typical MMF configuration (128 km domain size and 4 km model grid spacing).