A Review on Machine Learning Algorithms for Dust Aerosol Detection using Satellite Data
Nurul Rafi, Pablo Rivas
TL;DR
The paper addresses the problem of detecting and modeling dust aerosols from satellite data using machine learning, tracing a historical shift from simple spectral-band indices to sophisticated data-driven models. It surveys a broad set of sensors (MODIS, CALIPSO, VIIRS, CALIOP, MERIS, AVHRR, OMI) and physical dust indices (e.g., $BTD_{3-11}$, $NDDI$), then canvasses ML families from SVM to CNNs and ensemble methods. Key findings show that ML models generally improve dust detection accuracy and enable tasks beyond simple detection, while challenges such as explainability and scalability persist. The work provides a practical roadmap for future research, highlighting attention-based temporal models, hyperspectral CNNs, and hybrid semi-supervised approaches as promising directions for advancing dust aerosol remote sensing.
Abstract
Dust storms are associated with certain respiratory illnesses across different areas in the world. Researchers have devoted time and resources to study the elements surrounding dust storm phenomena. This paper reviews the efforts of those who have investigated dust aerosols using sensors onboard of satellites using machine learning-based approaches. We have reviewed the most common issues revolving dust aerosol modeling using different datasets and different sensors from a historical perspective. Our findings suggest that multi-spectral approaches based on linear and non-linear combinations of spectral bands are some of the most successful for visualization and quantitative analysis; however, when researchers have leveraged machine learning, performance has been improved and new opportunities to solve unique problems arise.