- Related Research Areas
- Atmospheric Composition
The overall goal of this proposal is to improve cloud and radiation parameterizations in the GEOS-5 earth system model and data assimilation system. This will be achieved by developing novel cloud assimilation techniques coupled to a statistically consistent cloud scheme that is able to properly account for subgrid variability at the cloud scales. This parameterizations is specifically designed to extract the maximum information content of high resolution pixel data available with the NASA A-Train instruments. Extensive use of MODIS, OMI, CERES, AMSR-E, CloudSat and CALIPSO data will be made in conjunction with stochastic cloud generators to improve radiative transfer calculations in GEOS-5. The main objectives of this proposal are: 1.Develop an improved GEOS-5 cloud parameterization. The new scheme will model the distribution of total water in each GEOS-5 layer and the vertical correlations of those distributions. Specifically, we will employ so-called copula functions to model the complex vertical correlations in the moisture field while retaining full control over skewed horizontal moisture distributions in each model layer. Compared to simpler cloud generation approaches, our method provides a much improved treatment of cloud fraction and the cloud overlap problem (both geometrical overlap and vertical condensate correlations), leading to a more accurate cloudy radiative transfer calculation in GEOS-5. The method is also ideally suited to generation of cloudy subcolumn realizations for satellite instrument simulators. 2.Constrain the new cloud parameterization by assimilation of A-Train cloud data. The richness of cloud related measurements provided by MODIS, AMSR-E, OMI, CloudSat and CALIPSO high resolution cloud data will be fully explored to constrain this new PDF-based cloud scheme. This work will expand upon the Norris and da Silva (2007) parameter estimation approach to cloud assimilation and will take advantage of the high order subgrid-scale statistical information provided by the aforementioned A-Train Level-2 data. 3.Improve the GEOS-5 radiative transfer (RT) by means of a stochastic cloud generator. The new statistical cloud scheme very naturally provides for improved cloudy radiative transfer via a stochastic cloud generator. This generator produces an ensemble of subcolumns that collectively represents the modeled horizontal and vertical sub-gridscale variability of moisture in GEOS-5. To efficiently perform short- and longwave radiative transfer on this ensemble, we will implement the Monte-Carlo Independent Column Approximation (McICA) method of Pincus et al. (2003) on existing GEOS-5 RT algorithms. The more accurate RRTMg radiative transfer codes will also be implemented and evaluated in GEOS-5 in both their standard and McICA incarnations. We will produce a publicly available comprehensive targeted reanalysis of the CloudSat period, with extensive validation by non-assimilated data products. This proposal makes a direct contribution to NASA's Strategic Sub-goal 3A: Study Earth from space to advance scientific understanding and meet societal needs.
Project PI: Arlindo da Silva/NASA Goddard Space Flight Center
Phone: (301) 614-6189
Fax: (301) 614-6307
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 3 years, 8 months ago
Visit our help center