Evaluation of two Eta Model versions for weather forecast over South America

Abstract

A comparison of performance of two versions of the Eta/CPTEC model is presented. The new version is an update of an earlier operational one and includes representation of soil and vegetation types. The soil/vegetation model contains two underground layers and a canopy layer. Evaluation was carried out by comparing the mean and root mean square errors of several variables for an ensemble of 15 meteorological situations, using the NCEP analyses, upper air soundings and precipitation data over South America. Precipitation forecasts were evaluated by the equitable threat score (ETS) and the bias score (BIAS).The mean errors from both versions show similarities during the first 24 hours of forecast, but surface temperature is more accurately predicted by the updated model. After 48 hours, temperature and humidity forecasts show better skill in the new version too. Over subtropical South America differences are more evident in temperature and humidity, less so for geopotential heights and practically nonexistent for horizontal winds. The precipitation forecasts for the updated version have equitable threat scores slightly higher and similar bias scores during the first 24 hours. After 48 hours this version tends to overestimate the rainfall, while its spatial distribution remains unaffected.

Keywords

• Meteorology
• regional modeling
• performance
• surface processes
• South America