Investigation of drought in terms of length, intensity and frequency in the Upper Sakarya Basin

Drought, one of the most costly natural disasters worldwide, comes in many types such as meteorological, agricultural and hydrological drought. Therefore, there is a need for improved methods in drought monitoring and decision-making. Most drought indices take into account one or two driving factors. This is not sufficient due to the complex nature of drought. Multivariate drought indices have been developed that combine the advantages and reduce the disadvantages of existing approaches. In this study, SPEI (StandardizedPrecipitationEvapotranspiration Index) and PDSI (PalmerDroughtSeverity Index), which take into account a large number of variables, were used to evaluate the meteorological, agricultural and hydrological drought in the study area of Upper Sakarya Basin, which is one of the hotspots of the ecosystem in terms of industry, agriculture, culture and landscape, including agricultural, forest, pasture and settlement land classes in Turkey. The region, which has three different characteristics (BSk: Semiaridcold, Csa: Drysummer-hot summer and H: Unclassifiedhighland), was classified based on Standardized Precipitation Evapotranspiration Index (SPEI) and Palmer Drought Index (PDSI) data. Precipitation, evapotranspiration and runoff data for the period 2003-2021 were obtained from SPEI Global Drought Monitor and ERA5 databases. R packages that offer a wide range of applications for drought modeling and assessment based on multivariate drought indices were used in the study. Both SPEI and PDSI 1-3-6-12 (soil moisture, surface hydrology and agricultural and hydrological perspectives) at 4 different time scales showed acceptable performances in multivariate drought prediction. SPEI values obtained as a result of modeling at all four time scales (1, 3, 6 and 12 months) were compared with PDSI values and similarities and differences were discussed. All drought data were also evaluated in terms of frequency, magnitude and length. ERA5 database contains global scale data sets with a standard geometric resolution of 0.50o×0.50o since 1950. Therefore, the Upper Sakarya Basin, where the study will be carried out, was evaluated by dividing the study area into 39 grid regions at 0.50o×0.50o scale. Thus, it was tried to obtain data that can be evaluated in four time scales for each section and the whole area in the basin divided into 39 grid regions.