Assessment of Drought Severity Based on Remote Sensing Using a Multi-scale Intelligent Method (Case Study: Northwest Iran)

Drought as a natural disaster is one of the ecological, hydrological, agricultural and economic concerns of humans. In this study, a multi-scale intelligent method was used to assess drought severity and drought-prone areas in northwest Iran for the years 1983-2020. The modeling process utilized both remote sensing and ground-based datasets. To analyze drought conditions. drought time series were first decomposed into several subseries using the Variational Mode Decomposition (VMD) method. Then, the most effective subseries were selected based on their energy values and the selected area was zoned. In the next step, the mean energy values of each station were used as input of the K-means clustering technique. Results proved the proposed multiscale method's appropriate efficiency in effectively diagnosing drought severity. Results showed that the southeastern part suffers from severe drought, while drought in the southwestern and northwestern parts of the study area is milder. The lowest energy values were  obtained for the central regions, where drier areas were present. It was found that there is an inverse relationship between the drought index and energy values. Also, based on the results obtained from clustering, it was revealed the similarity of intra-cluster hydrological uncertainty, although some rain-gauge stations in some clusters are located at a great distance from each other. In other words, the basis of clustering was not necessarily the proximity of the stations.