Share Email Print
cover

Proceedings Paper

Using IR satellite data to improve the initialization of a mesoscale model
Author(s): B. Codina; A. Sairouni; Josep Vidal; Angel Redano
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

An empirical technique to blend satellite imagery into a conventional meteorological analysis has been developed in order to enhance the 3D humidity field in a numerical weather prediction model. Temperatures retrieved from IR Meteosat images are used to locate and grossly characterize clouds, and this information further employed to dry and moisten individual atmospheric cells from a previous conventional analysis. To verify the improvements attained with this technique, we have compared a set of 260 forecasts obtained with a conventional initialization of the mesoscale model MASS, and those from the enhanced initialized. Standard indexes, such as mean error, k root mean squared error and S1, have been used to objectively evaluate the quality of the forecasts over the Southwestern European region. It turns out that the inclusion of satellite imagery in the initial analysis leads to improvements up to 30 percent in more than 200 out of the 260 days considered over the typical meteorological fields. Although for several reasons no attempt has been made to compare complex fields such as precipitations, it is obvious that an improved depiction of the driving fields might result in better weather forecasts.

Paper Details

Date Published: 14 December 1998
PDF: 6 pages
Proc. SPIE 3495, Satellite Remote Sensing of Clouds and the Atmosphere III, (14 December 1998); doi: 10.1117/12.332678
Show Author Affiliations
B. Codina, Univ. de Barcelona (Spain)
A. Sairouni, Univ. de Barcelona (Spain)
Josep Vidal, Univ. de Barcelona (Spain)
Angel Redano, Univ. de Barcelona (Spain)


Published in SPIE Proceedings Vol. 3495:
Satellite Remote Sensing of Clouds and the Atmosphere III
Jaqueline E. Russell, Editor(s)

© SPIE. Terms of Use
Back to Top