dust driver emission

This dataset was compiled to support the machine learning-based prediction of dust emission hotspot activation in the Tigris-Euphrates Basin for the period of 2000-2021. It contains two main components: the dust hotspot locations and the corresponding environmental driver data. The target data includes the spatiotemporal coordinates of 11,618 dust hotspot events, which were identified through the visual interpretation of MODIS satellite imagery, providing a binary 'active' or 'non-active' status for each location on a monthly basis. For each hotspot location, the dataset provides a corresponding time-series matrix of 11 key environmental drivers, including meteorological data from TerraClimate (precipitation, wind speed, soil moisture, temperature, NDVI), and static geophysical properties (soil texture, soil thickness, altitude, and slope). All dynamic variables are provided at a monthly temporal resolution, and all features have been harmonized to a 1 km spatial grid to create a consistent modeling framework.