Spatial variability of meteorological observations and impacts on regional estimates of soybean grain productivity

Abstract

Brazil requires a fully representative weather network station; it is common to use data observed in locations distant from the region of interest. However, few studies have evaluated the efficiency and precision associated with the use of climate data, either estimated or interpolated, from stations far from the agricultural area of interest. Hence, this study aimed to demonstrate the impacts of spatial variability of the main meteorological elements on the regional estimate of soybean productivity. Regression analysis was used to compare data recorded at three weather stations located throughout Londrina, PR, Brazil. The water balance of the soybean crop was calculated at 10-day periods and grain productivity losses estimated using the Agro-Ecological Zones (AEZ) methodology. Temperatures at the three locations were similar, while the relative air humidity, and particularly, the rainfall data, were less correlated. A high degree of caution is recommended in the use and choice of a single weather station to represent a municipality or region, particularly in countries, such as Brazil, with multiple regions of agricultural and environmental importance. Models and crop season estimates that do not consider such a recommendation are vulnerable to errors in their forecasts. The volumetric and temporal variability in the spatial rainfall distribution resulted in soybean yield discrepancies, estimated at the municipal level. The consistency of the data series, the location of weather stations and their distance to the location of interest determine the ability of crop models to accurately estimate soybean production based on meteorological data, particularly the rainfall data. This study contributes to future regional research using climate data, and highlights the importance of a weather station network throughout Brazil, demonstrating the urgent need to increase the number of weather stations, particularly for recording rainfall data. Ferreira, R. C. et al escassos estudos que avaliem a eficácia e precisão da utilização de dados climáticos estimados ou interpolados a partir de estações distantes da área agrícola de interesse. Assim, este estudo teve como objetivo demonstrar os impactos da variabilidade espacial dos principais elementos meteorológicos sobre a estimativa regional da produtividade de grãos de soja. Utilizaram-se dados observados em três estações meteorológicas em diferentes locais de Londrina, comparados por meio de análise de regressão. Calculou-se o balanço hídrico decendial para soja e estimaram-se as perdas de produtividade de grãos pelo método Zona Agroecológica. As temperaturas nos diferentes locais apresentaram semelhanças, enquanto a umidade relativa do ar e, principalmente, precipitação pluvial foram mais discrepantes. Recomenda-se muita cautela no uso e na escolha de uma única estação meteorológica para representar um município ou região, como acontece em várias regiões de importância agrícola e ambiental no Brasil. Modelos e resultados de estimativas de safras que não consideram tal...

Effect of Water Stress on the Development of Soybean Crop

“…According to Ferreira et al [19] the consistency of the data, the location and the distance from the meteorological stations to the place of interest are determining factors of precision of grain yield estimates based on meteorological data, mainly precipitation data. In their studies, soybean water balance was calculated with data recorded in three meteorological stations where they proved variability in rainfall distribution, which resulted in soybean yield discrepancies, estimated at the regional level.…”