Comparação entre dados meteorológicos estimados e observados para utilização em modelos de estimativa da produtividade agrícola Marcelo

Abstract

Due to the importance of soybean crop for the Brazilian economy as well as for the Rio Grande do Sul State (RS) and the consequent need to monitor the crop yield conditions, this study aims to compare meteorological data generated by the regional weather forecast model ETA and the Tropical Rainfall Measuring Mission (TRMM) system versus observed data by conventional weather stations regarding their use in agronomic models in order to monitor and estimate soybean crop yield in RS. It was used original or derived data from rainfall, average air temperature, downward short wave radiation, average relative air humidity, and 2-m average wind speed observed by weather stations from INMET as reference data to compared with those estimated by the TRMM system (rainfall) and the ETA model (for other variables). It was used three crop years (2006/07 to 2008/09) for comparison and generation of the adjustment models and the three subsequent crop years (2009/10 to 2010/11) for validation. Results showed high correlation (r) between rainfall data retrieved from TRMM and INMET for most crop years and the estimate accuracy was low altered after the adjustment by the mathematical model. In general, for average air temperature, downward short wave radiation, and average relative air humidity, results showed moderate correlation between ETA and INMET data. Moreover, the majority of the ETA data was improved when the mathematical models were applied. For 2-m average wind speed, results showed very low correlation between ETA and INMET data.