IMPROVED CLOUD-RADIATION and HYDROLOGIC CYCLE PARAMETERIZATIONS for MODELING and PREDICTING CLIMATE VARIABILITY

Richard Somerville (SIO)

Link to NOAA Strategic Plan: NOAA's Mission Goal 2: Understand Climate Variability and Change to Enhance Society's Ability to Plan and Respond

RESEARCH OBJECTIVES AND SPECIFIC PLANS TO ACHIEVE THEM

The major focus of this project is the development of improved parameterizations of cloud-radiation effects and related processes, using the diagnostic SCM to make direct comparisons of results from parameterizations with satellite remote sensing observations. Our group at Scripps and our collaborators base these parameterizations on physically comprehensive detailed cloud microphysics modules derived from earlier and ongoing work. These parameterizations will then be incorporated into modern atmospheric GCMs and the results evaluated against satellite-based climatologies of cloud-radiation variables.

RESEARCH ACCOMPLISHMENTS

The control version of the SCM incorporates the prognostic cloud scheme of Tiedtke (1993) that includes precipitation formation based on Sundqvist et al. (1989). The prognostic cloud scheme developed in the SCM was incorporated into the Experimental Climate Prediction Center Global Spectral Model (ECPC-GSM). Results from the ECPC-GSM with the new prognostic cloud scheme were encouraging as the cloud fraction and cloud water/ice contents were generally more realistic than either the original ECPC-GSM control cloud parameterization or other new cloud parameterizations also tested. The results from the ECPC-GSM as well as results from tests with other GCMs and SCM work suggested deficiencies in the precipitation physics may be leading to some errors in the cloud water/ice path and cloud amount.

A more physically realistic precipitation parameterization was developed for the SCM. This new scheme uses an auto-conversion formulation similar to that of Manton-Cotton (1977) and includes separate equations for each microphysical process such as accretion, collection and flux divergence of falling ice particles. The SCM now incorporates completely independent equations for liquid and ice. SCM results using this new precipitation scheme show significant improvement in the profiles of cloud water and ice when tested against measurements from the DOE Atmospheric Radiation Measurement (ARM) Program.

In collaboration with Drs. Masao Kanamitsu and Akihiko Shimpo, our improved parameterizations were incorporated into the version of the ECPC-G-RSM model running at Scripps under the direction of Dr. Kanamitsu. The G-RSM was forced with observed SST and seasonal mean fields of cloud-radiation variables evaluated. Our parameterizations produced modest improvements in the geographical distribution of cloudiness and precipitation compared to other schemes tested in the model. The results from this study also showed that the results are very sensitive to parameters in the cloud schemes.

We also incorporated our improved cloud parameterizations into the NCAR CAM3 model and produced 10-year runs at T-31 resolution. Compared to the control version of CAM3, our parameterizations produced more realistic values of cloud water amount and cloud amount in the mid-latitudes, while cloud amount in the tropics were over-estimated. The model results using our parameterizations also reduced an upper tropospheric mid-latitude cold bias and a bias in the mid-latitude zonal wind that were present in the control version of CAM3.

The next phase of this research will explore how well these results generalize to other models, and there we will be focused on the NOAA global forecast system. We believe that parameterizations developed and evaluated in climate GCMs may have positive impacts in NWP models and vice versa.

Fig. 1 Annual averaged column-integrated cloud water path differences (model-observation) from control case (top panel) and from experiment case (bottom panel) of a 10-year integration of CAM3 at T31 resolution. Observations are taken from the NASA Water Vapor Project (NVAP). The experiment run utilizes our improved cloud parameterizations and results in a large improvement in the horizontal distribution of cloud water. We believe that experiments such as this one utilizing climate models to develop and test parameterizations can have positive impacts when these schemes are later incorporated into NWP models. (Units are g m-2)