Rayegani, B; Barati, S; Goshtasb, H; Gachpaz, S; Ramezani, J; Sarkheil, H (2020). Sand and dust storm sources identification: A remote sensing approach. ECOLOGICAL INDICATORS, 112, 106099.

In recent years, the sand and dust storm (SDS) events have become one of the most critical environmental challenges around the world. Identification of the sources not only helps to monitor and predict the processes of dust storms, but also it contributes to the reduction of its negative impacts and management of this phenomenon in a better manner. Identifying the sources based on field observation is impossible due to the vast extent of these lands, as well as the limitation of access to most of their areas, so remote sensing data can be used as a reliable alternative. The primary objective of this study is developing a comprehensive approach for sand and dust storm source identification and surveying their changing trend during a specific period via remotely sensed data. For this purpose, to generate wind erosion sensitivity maps based on the vegetation cover, soil moisture and land cover, Landsat8 data from 2013 to 2015 were acquired. After undergoing pre-processing on Landsat data with the help of spectral vegetation indexes and classification, wind erosion sensitivity maps were obtained for vegetation, soil moisture, and land use. To obtain a potential dust source map, the previously generated maps were integrated with geology and soil roughness information through multi-criteria evaluation to produce a wind erosion sensitivity map. Then, the synoptic data and air-quality information data were collected and using statistical analysis and MODIS data, local dust and sand events were identified, which were then validated based on HYSPLIT air flow simulation model to ensure that airflow trajectory and erodible lands are in physical contact. Based on the contact areas between airflow and the terrain besides applying non-erodible masks on them, the wind erosion risk was mapped. Next, the wind erosion risk map and the wind erosion sensitivity map via fuzzy multi-criteria assessment through linear weighted method were combined, then based on stratified-random sampling, probable SDS sources were identified. To validate the identified SDS sources and survey their trend, time series and synoptic data were utilized and vegetation cover, soil moisture and land surface temperature (LST) trend in specified areas for 15 years were monitored. Validation results showed high accuracy in identifying the areas; moreover, they confirmed the significant decrease in the vegetation cover, soil moisture and LST in the SDS sources during the study period (the last 15 years). Our results showed the high capability of time series analysis of remotely sensed data. LST as a climatic parameter has a crucial role to identify and validate the SDS sources. For instance, in areas with high SDS frequency, a significant decrease in LST is observable and confirmed by AOD analysis results. In this research, we integrated the most critical data that can be effective in identifying dust sources. To validate these sources, we used time series satellite data to measure the power of these data. Finally, we proposed the whole procedure as a complete process. Hence, the applied method in this research can be used as a comprehensive approach for future studies in SDS source identification by remotely sensed data.