An Approach for Using Remote Sensing Products and Ground Observations in the Evaluation of a Numerical Weather Prediction Model

Abstract

A common problem in Air Quality Modeling (AQM) is the lack of meteorological observations in the area of interest for the initialization of Numerical Weather Prediction (NWP) models or evaluation of their results. Even if a significant number of weather stations are available, they cannot provide information above the surface level. Satellite atmospheric products constitute a unique source of information as they can give measurements at every point of the planet, and vertical profiles of some meteorological variables. However, this information should be validated if operational use is intended in air quality studies. This work develops a methodology for spatial and temporal forecast verification using three different sources of information regularly available in Mendoza: MODIS atmospheric products, radiosounding data and radiometer measurements of water vapor. We present comparisons and statistical analysis between Moderate Resolution Imaging Spectroradiometer (MODIS) temperature and dew point profiles versus traditional upper air profiles obtained by soundings and total water vapor content obtained with radiometer. We developed statistical indicators of information quality, as a guide for operational use of satellite atmospheric products for NWP model validation.