Evaluation of Cloud Parameterizations in a High Resolution Atmospheric General Circulation Model Using ARM Data

Abstract

Typical state of the art atmospheric general circulation models used in climate change studies have horizontal resolution of approximately 300 km. As computing power increases, many climate modeling groups are working toward enhancing the resolution of global models. An important issue that arises when resolution of a model is changed is whether cloud and convective parameterizations, which were developed for use at coarser resolutions, will need to be reformulated or re-tuned. We propose to investigate this issue and specifically cloud statistics using ARM data. The data streams produced by highly instrumented sections of Cloud and Radiation Testbeds (CART) of ARM program will provide a significant aid in the evaluation of cloud and convection parameterization in high-resolution models. Recently, we have performed multiyear global-climate simulations at T170 and T239 resolutions, corresponding to grid cell sizes of 0.7{sup 0} and 0.5{sup 0} respectively, using the NCAR Community Climate Model. We have also a performed climate change simulation at T170. On the scales of a T42 grid cell (300 km) and larger, nearly all quantities we examined in T170 simulation agree better with observations in terms of spatial patterns than do results in a comparable simulation at T42. Increasing the resolution to T239more » brings significant further improvement. At T239, the high-resolution model grid cells approach the dimensions of the highly instrumented sections of ARM Cloud and Radiation Testbed (CART) sites. We propose to form a cloud climatology using ARM data for its CART sites and evaluate cloud statistics of the NCAR Community Atmosphere Model (CAM) at higher resolutions over those sites using this ARM cloud climatology. We will then modify the physical parameterizations of CAM for better agreement with ARM data. We will work closely with NCAR in modifying the parameters in cloud and convection parameterizations for the high-resolution model. Our proposal to evaluate the cloud parameterization at high resolution has high relevance for the mission of ARM (to improve the treatment of clouds and radiation in the models) because the high-resolution model grid cells approach the dimensions of the highly instrumented sections of ARM CART sites.« less