12. Title: Physically-Consistent Database of MW-3D Profiles in La Plata Basin and Spain: setup for TRMM and application to GPM
PI: Francisco J. Tapiador
Budget: € 381,000 (approx)
Period: Mar 2008 –Mar 2010
Summary
A RM quasi-ensemble is a high-resolution limited
area model (LAM) with a single dynamical core and several
parameterizations accounting for microphysics, radiation, PBL and
surface processes that are unresolved even at the LAM high resolution.
Each realization of the RM for a given parameterization is called an
ensemble member. The dispersion of the ensemble members provides the
limits of the actual atmospheric state, with a non-linear combination
of the members being the most likely state, and the 2σ- spread of the
ensemble being the expected error.
The goal of this project is to setup a system
comprised by a Radiative Transfer calculation (RTC) module coupled with
a quasi-ensemble of Regional Models (RM). This system will be run on
Spain and La Plata Basin (LPB) to generate physically consistent
match-ups of MW radiances with raining and non-raining atmospheric
parameters.
The resulting database might be of the interest of
GPM algorithms developers in terms of analyzing algorithmic dependence
on the physical assumptions made. By using RT, we can calculate the
satellite-like expected radiances for each one of the RM ensemble
members. Each of those estimates will be a satellite simulation for the
atmospheric column under a given parameterization. The extensive
parameterizations will include cumulus parameterizations such as Grell
and Kain-Fritsch, and explicit moisture schemes such as Schultz and
Goddard microphysics.
Comparison of actual satellite measurements (TRMM
for LPB, TRMM and SSM/I for Spain) with the simulated fields and with
ground validation data will provide insight into suitable physical
parameterizations for the two study areas, thus improving precipitation
retrievals. Moreover, the Jacobian of the RT calculation will permit
using the satellite estimates in a data assimilation scheme working in
ensemble mode. Given the computation burden of ensemble calculations,
we will use the Barcelona Supercomputing Center facility to carry out
our calculations.