Drought analysis and its relation with wheat yield in Konya Altinova region with satellite-based indexes

In recent years, Konya, one of the most important cereal production centers in Türkiye, has experienced severe drought events, which have negatively impacted crop yields, particularly wheat. This dissertation aims to analyse drought in the Altınova region of Konya province through up-to-date approaches and to develop models that can be used to estimate the yield of typical wheat varieties using the results obtained. The drought trend from the past to the present was analysed, and short-, medium-, and long-term drought projections for the future were developed. The yield prediction models for the Kızıltan-91, Bayraktar-2000, and Bezostaya-1 varieties were generated using machine learning algorithms. According to the study findings, SPI analyses drought recurrence between 28-31% and SPEI between 33-34%, and there may be a hydrological drought trend on the SPEI-24-month time scale. According to Future projections, no global model or time scale indicates drought tendency in a short-term (2020-2040). It is concluded that combining SPI and SPEI for meteorological drought will allow for far more accurate assessments. Agricultural drought analyses were carried out using satellite-based drought indices (VCI, TCI, VHI and VSWI) for wheat growth phases. Regarding application and outcomes, the 'Vegetation Condition Index (VCI)' is deemed suitable for the region. According to the VCI, all varieties experienced agricultural drought at various times in 2001, 2004, 2007, 2008, 2012, 2014, 2016, 2017, 2018 and 2019. Finally, the study presented 21 yield prediction models for the Bayraktar variety, 12 yield prediction models for the Kızıltan variety, and 8 yield prediction models for the Bezostaya variety as alternatives, and the model performances (R2) ranged between 0.736-0.973, 0.733-0.956 and 0.686-0.874 respectively. It is concluded that the developed models could be implemented in the region.