Retrieval of Flux Properties Associated with Mesoscale Winds

Convective momentum properties have been studied for many years using in-situ or aircraft measurement, along with numerical simulation. New research involving TRMM has been conducted (Mecikalski 2003) showing that satellite remote sensing can be a useful tool in diagnosing the momentum properties and fluxes occurring in a convective regime. Although previous research has been to utilized radar data to obtain momentum flux properties (see Rabin et al. 1982; Angevine 1994), little or no work has been performed using satellite-derived winds to analyze properties such as momentum fluxes associated with a convective environment. Therefore, this project's goal is to determine whether satellite-derived winds can accurately depict mesoscale fluxes.

Using satellite-derived winds for mesoscale atmospheric motion vectors (MAMVs) as described by Bedka and Mecikalski (2005), one can obtain the actual winds occurring within a convective environment. Outflow boundaries and upper-tropospheric anvil outflow will produce “perturbation” winds on smaller scales that the MAMV algorithm can measure. Using a computed background wind field from either NWP or the actual satellite winds (smoothed/averaged over a larger region), horizontal perturbations, u' and v', associated with the mesoscale flow can be calculated. Ongoing study is toward determining the optimal method for estimating the background wind field, and performing simple comparisons to GOES 1 km (visible imagery) observed cumulus over uniform surfaces. As an initial test, analysis is being done over the Caribbean on one September day when many cumulus cloud features were present, henceforth allowing for the retrieval of many (ca. 50000) MAMVs used in this analysis.

To ensure that these fluxes are accurate, a cloud-resolving model will be run for the same time and geographical area. Assuming that the model is a form of “truth” in terms of physically reproducing the convective structures, a comparison will be made between the model's fluxes and the satellite-derived fluxes to see if the derived fluxes are accurate not only in magnitude but also in sign. Determining the accuracy of the derived fluxes will measure our ability to determine convective momentum fluxes.

Additional work will involve using several means of estimating perturbation vertical velocities, as produced within the observed clouds (leading ultimately to the observed u' and v' fields). Our initial effort will be to determine if w' may be attained based on convergence/divergence in the cloud-top and boundary layer respectively or through the use of Doppler radar. Once the vertical and horizontal perturbations are calculated, vertical convective momentum fluxes, (u'w') and (v'w'), can be computed.