- Related Research Areas
- Atmospheric Composition, Climate Variability & Change
In the same way in which line-by-line radiative transfer models have served as a benchmark for the evolution and validation of clear-sky radiation parameterization, a self-consistent, comprehensive, and validated data set of key cloud variables can greatly contribute to the development and validation of cloud parameterizations in large-scale weather and climate models. The intrinsically complicated multi-scale interaction between radiation, dynamics, and cloud microphysics requires such a "benchmark" dataset to have the capability to correctly represent relevant multi-scale processes and their couplings. The NASA EOS missions and ROSES MAP program have produced new resources for studying these multi-scale phenomena, i.e., the advance of cloud resolving models such as Goddard Cumulus Ensemble model (GCE), the unprecedented data availability from coordinated A-train observations, and the MERRA reanalysis effort with its emphasis on proper representation of the hydrological cycle. This proposal embraces the timely opportunity provided by these assets, and consists of an initial study toward the grand goal of assembling a "benchmark" dataset for evaluating cloud parameterizations. The premise of this proposal is that neither EOS observations nor cloud resolving models alone are sufficient to provide a reliable test-bed dataset. Instead, what is required is a synergy that brings together both components in the context of improved large-scale tendencies from MERRA. As an effort in line with this premise, we propose to explore optimal ways to use EOS observations (chiefly CloudSat, AIRS, and CERES radiative fluxes) to constrain GCE simulations forced by MERRA large-scale tendencies (a process we have termed "optimal blending") and to validate the utility of such an approach for producing a useful dataset for evaluating GCM cloud parameterizations. We will focus on the South China Sea and ocean regions adjacent to the Maritime Continent and north Australia, where strong constraints from in-situ observations render a highly confident MERRA reanalysis, the GCE model has proven skill in simulating major cloud features, and reliable retrievals can be obtained from EOS instruments. The primary tasks to be accomplished under this proposal are: (1) Compositing EOS observations in conjunction with the ISCCP deep convection tracking database to capture domain-averaged features of different stages in the lifecycle of cloud systems and their interaction with the large-scale circulation; (2) Carrying out GCE simulations driven by MERRA large-scale tendencies, evaluating the simulations against EOS composites, and understanding the consistencies and conflicts between observations and simulations; (3) applying a Markov chain Monte Carlo technique to evaluate the GCE cloud microphysical scheme and obtain a set of optimal parameters for each individual stage of the cloud systems; (4) Validating the parameters derived in task (3) by further comparing simulations with independent EOS observations (e.g. precipitation from AMSR-E and TRMM) and other appropriate observations. The proposed work aims to improve our understanding of cloud multi-scale processes via orchestrated usage of EOS observations, GCE models, and MERRA reanalysis. Such improvement is a crucial step towards improved numerical simulation of future climate change, and is directly related to NASA's Science Question "How will the Earth system change in the future" and NASA's research objective 3A.2 "Enable improved predicative capability for weather and extreme weather events".
Project PI : Xianglei Huang/University of Michigan
1541A Space Research Building 2455 Hayward Street, Ann Arbor, MI 48109-2143
Phone: (734) 936-0491
Fax: (734) 936-0503
Email: xianglei @umich.edu
Nothing to see here at the moment. Check back later.
Log in to start a discussion.
- Only approved users can join
- Anybody can view this project
- Any registered users can leave comments
- Anybody can view comments
- Joined 4 years, 7 months ago
Visit our help center