Remote sensing data for drought stress and croplands productivity assessment in Kherson region

Formulation of the problem. Remote sensing data might be used for indirect assessment of croplands conditions and drought stress through the calculation of specific vegetation indices, such as vegetation health index (VHI), agriculture stress index (ASI), and drought intensity or weighted mean vegetation health index (WMVHI). However, the accuracy of these indices is not clear for some territories. For example, the South of Ukraine is a zone of risky agriculture, because of low natural moisture supply and high evapotranspiration. Moisture supply is the main limiting factor for sustainable crop production in this region. The goals of this study were: 1) to assess the reliability of the mentioned vegetation indices in drought assessment through the direct comparison with the UNEP aridity index; 2) to find out whether remote sensing drought indicators could be used for the yield prediction of major crops on the regional scale. Methods. The study was conducted for Kherson region of Ukraine, as it is one of the most arid regions of the country with very high drought risks. The data on average weighted annual VHI, ASI, and WMVHI for the period 1984-2022 (Season 1) were collected and generalized from the FAO Earth Observation services. UNEP aridity index was calculated using the data from Kherson regional hydrometeorological center. Correlation and linear regression analysis were performed using common statistical methodology. Results. As a result, it was found that 1) all the studied remote sensing drought indicators demonstrate poor correlation with the aridity index, therefore, they should not be used to determine meteorological drought in the region; 2) all the studied remote sensing indices, especially VHI, demonstrate moderate-to-strong correlation with the yields of certain crops, cultivated in Kherson region (R=0.54-0.86), and could be used for the yield prediction; 3) the aridity index have poor relation to the yields of major crops, cultivated in the studied area; 4) VHI-based linear regression models for the crops’ yields prediction are reliable and reasonable for scientific and practical use just for cereal crops, and are much less accurate for grain corn and sunflower; 5) based on the study findings, it could be concluded that aridity index provides pure climatological characteristics of the region, while the studied vegetation indices are mainly focused on the level of drought stress that impacts crops during the growing season. Scientific novelty and practical significance. The article provides novel insights on the implementation of remote sensing data in drought risks assessment in crop production, and their utilization for the purpose of croplands productivity prediction. The study has theoretical and practical importance for current agriculture, and the findings could be used both in scientific, educational, and practical purposes.