New approaches toward the evaluation of convective momentum fluxes [i.e. (u'w' and v'w')] from satellite data are developed. Data this study uses are Tropical Rainfall Measuring Mission (TRMM) precipitation radar (PR; TRMM product "1C21") and microwave imager data (TMI; TRMM product "2B31"), GOES visible and infrared cloud-motion winds, and near-surface scatterometer winds from QuikSCAT. The primary goal is to develop techniques that allow for robust estimates of convective momentum fluxes on storm- to meso-scales. Momentum flux signatures that our methods key in on are precipitation patterns (rainfall signatures; Mecikalski 2003) within TRMM data, and highly divergent/ convergent surface and upper tropospheric flows near convective clouds as observed by cloud-motion and scatterometer winds. Use of rainfall accumulation/rate estimates from TRMM (products "2A12" and "2A25") also provides our methods important information that is used to analytically derive vertical velocities within convection (i.e. w').

As a means of validating these estimates, cloud resolving model simulations (at 1-2 km resolution) over the Kwajalein Atoll Experiment (KWAJEX), the TExas-FLorida UNderflights-B (TEFLUN-B), and South China Sea Monsoon Experiment (SCSMEX) domains are used to estimate "truth" momentum budgets for the same regions our satellite data analyses are performed. Error estimates between the simulated and estimated momentum fluxes are done to assess accuracy (in terms of flux sign and magnitude) and improve our satellite-based methods. Evaluation of the convective momentum budgets of the KWAJEX, TEFLUN-B and SCSMEX regions are an outcome of the cloud resolving modeling component of this project.

An additional aspect of this work couples the satellite flux estimates to a numerical model (run in non-cloud resolving mode at >15 km resolution) through the use of a cumulus parameterization. This represents a preliminary approach to the assimilation of these data into numerical models. Specifically, use and modification of the Emanuel (1991) and Wu and Wanai (1994) parameterization schemes allow for the direct assimilation of the above satellite information into numerical simulations by relating these information to model variables (tendencies of u, v and w winds) via the methods developed above. Through assimilation, the importance of the satellite-based momentum flux retrievals are evaluated through the measured impact on large scale flows (when comparison to control simulations are made).