ВИВЧЕННЯ ВПЛИВУ МЕТЕОРОЛОГІЧНИХ УМОВ НА ВЕЛИЧИНУ НОРМАЛІЗОВАНОГО ДИФЕРЕНЦІЙНОГО ІНДЕКСУ НА ПОСІВАХ СІЛЬСЬКОГОСПОДАРСЬКИХ КУЛЬТУР У ХЕРСОНСЬКІЙ ОБЛАСТІ

Abstract

The article is devoted to the issue of establishing a correlation-regression relationship between meteorological parameters and the values of NDVI in crops such as sunflower, soybean, grain corn, and rapeseed. The purpose of the scientific work was to study the influence of meteorological conditions, namely air temperature, and precipitation, on the value of the satellite Normalized Differential Vegetation Index in the crops grown in the Kherson region, and to establish a correlation-regression relationship between meteorological parameters and spectral characteristics of the crops. Methods. The research was carried out on the territory of the Kherson region. Meteorological data were retrieved from the archives of the Kherson Regional Hydrometeorological Center and the data of the MeteoBlue open meteorological hub. Data on the values of NDVI were adapted and calculated using GIMMS Global Agricultural Monitoring. The data on the values of NDVI were calculated for each studied crop using their crop mask. The study was carried out for 2017 and 2020 for corn, 2017 for soybeans, 2021 and 2022 for sunflower and rapeseed, respectively. Correlation-regression analysis was performed according to standard methodology with a calculation of Pearson correlation coefficients, coefficients of determination, and coefficients of linear regression models. All calculations were performed at P<0.05 in the statistical package BioStat v.7. Results. It was established that the maximum closeness of the relationship is for the values of NDVI on sunflower and soybean crops with the air temperature (r=0.84 and 0.80, respectively). The closeness of the relationship between the NDVI values and the amount of precipitation was much smaller for all studied crops and can be considered significant only for soybean (r=0.45) and rapeseed (r=0.58). The calculation of regression coefficients and corresponding statistics proved that reliable models can be obtained for all studied crops in the pair “NDVI – air temperature”, and in the pair “NDVI – precipitation” a reliable model was obtained only for soybean. The best accuracy parameters are observed for the models in the pair “NDVI – air temperature” for such crops as sunflower, soybean, and corn – the relative error was 10.98, 19.97 and 22.15%, respectively. As for the model in the pair “NDVI – precipitation” for soybeans, the relative error of 52.26% indicates the limited possibilities of its theoretical use and absolute unacceptability for practical purposes. Conclusions. The study’s results testify to a high closeness and reliability of the correlation relationship between the values of satellite NDVI and air temperature in all the studied crops. The closeness of the relationship between NDVI and precipitation is average or low, and the construction of a reliable and accurate regression mathematical model for the pair “NDVI – precipitation” is impossible. In the future, the study will continue with the expansion of the geographical component to the entire zone of the Southern Steppe of Ukraine to establish the characteristics of the relationships between the NDVI and the parameters of the weather conditions of the agroecological zone. Keywords: precipitation, correlation, regression, satellite monitoring, air temperature.