DRASTIC is a model used to reliably assess the pollution potential of aquifers at global location. Despite various criticisms of this model or its optimization, sometimes the territorial or geological conditions still must rely on this model approach based on its simplicity. The existence of sinkholes in the alluvial aquifer with an underlying karst aquifer is a clear example of the need to change or optimize the DRASTIC model to improve accuracy of prediction. In this study, the DRASTIC model was changed by adding two factors, the distance to a sinkhole and the sinkhole catchment factors. The modified DRASTIC model was then used to evaluate the pollution potential of Abarkouh aquifer in Yazd province, Central Iran. The intrinsic vulnerability index of Abarkouh aquifer was calculated by the weighted sum method in GIS. The resulting numerical values ranged between 62 and 176 for full study area. Without adding the two layers related to Sin-DRASTIC, the generic DRASTIC Index map showed summed values ranging between 54 and 130. The final Sin-DRASTIC map was divided it into 4 orders very low (<106; 722 km2, 78%), low (107-133; 186 km2, 20%), low to moderate (133-160; 15.3 km2, 1.7%), and moderate vulnerability (161-176; 1.7 km2, 0.3%) vulnerability. Both of the created maps showed high Area Under the Curve values at 76% and 78% for the generic DRASTIC and Sin-DRASTIC respectively. However, the area of the aquifer containing the sinkholes showed low and very low zones of contamination vulnerability on the generic DRASTIC map. The sensitivity analysis showed that the aquifer vulnerability to pollution is mainly controlled by the vadose zone impacts and the sinkhole catchment factor. Therefore, it is concluded that for geologic settings containing sinkholes that are open or contain higher vertical hydraulic conductivity sediment compared to the general matrix of the unconfined aquifer, the Sin-DRASTIC model method yields a superior prediction.
