1. Machine learning-based prediction of sand and dust storm sources in arid Central Asia
- Author
-
Wei Wang, Alim Samat, Jilili Abuduwaili, Philippe De Maeyer, Tim Van de Voorde, Geography, and Cartography and Geographical Information Science
- Subjects
remote sensing ,event scale ,google earthengine (GEE) ,General Earth and Planetary Sciences ,Susceptibility mapping ,Software ,Computer Science Applications - Abstract
With the emergence of multisource data and the development of cloudcomputing platforms, accurate prediction of event-scale dust sourceregions based on machine learning (ML) methods should beconsidered, especially accounting for the temporal variability in sampleand predictor variables. Arid Central Asia (ACA) is recognized as one ofthe world’s primary potential sand and dust storm (SDS) sources. In thisstudy, based on the Google Earth Engine (GEE) platform, four MLmethods were used for SDS source prediction in ACA. Fourteenmeteorological and terrestrial factors were selected as influencingfactors controlling SDS source susceptibility and applied in themodeling process. Generally, the results revealed that the random forest(RF) algorithm performed best, followed by the gradient boosting tree(GBT), maximum entropy (MaxEnt) model and support vector machine(SVM). The Gini impurity index results of the RF model indicated thatthe wind speed played the most important role in SDS sourceprediction, followed by the normalized difference vegetation index(NDVI). This study could facilitate the development of programs toreduce SDS risks in arid and semiarid regions, particularly in ACA
- Published
- 2023
- Full Text
- View/download PDF