Fine-grained Surface Classification and Long-term Mapping of Dust Source Areas in Sino-Mongolian Gobi Based on Dust Events and Multi-source Remote Sensing
Acta Interdisciplinary Science · 2026
DOI: 10.48014/ais.20260706001 ↗Kurz gefasst
Fachartikel in der Zeitschrift Acta Interdisciplinary Science, 2026. Verfasst von Zirui ZHANG und 6 weiteren Autor:innen.
Abstract (Originalsprache)
The Sino-Mongolian Gobi is the most important dust source area in East Asia, and its dust activities and land cover changes directly affect the ecological security and air quality of northern China. Taking 2000-2025 as the study period, this study uses multi-source remote sensing data (Landsat, MODIS, and Sentinel- 2A) , meteorological data, and the backward trajectory simulation method to conduct research on dust event backtracking, source area identification, fine-grained land surface classification, and long-term change analysis. Considering the land surface characteristics of the study area, we constructed a dust-sensitive land surface classification system and established a high-dimensional multi-source remote sensing feature set. Under the same sample and feature conditions, we developed and compared U-Net, Random Forest, and Support Vector Machine classification models. The results show that the U-Net model achieved the best classification performance, with an overall accuracy of 93. 83% and a Kappa coefficient of 0. 92. The U-Net model was therefore used for accurate source area identification and long-term mapping. The results indicate that 90. 5% of dust events in the study area occurred in spring. Over the past two decades, mobile sandy land and saline-alkali land have expanded continuously, semi-fixed sandy land has generally decreased, and frequent bidirectional conversions have occurred between dry riverbeds and Gobi. This study can provide a reference for dust prevention and control and desertification management in arid regions.
Zitieren
APA
ZHANG, Z., SONG, D., SUMIYA, E., WANG, B., YONG, M., BAO, Y., & XU, R. (2026). Fine-grained Surface Classification and Long-term Mapping of Dust Source Areas in Sino-Mongolian Gobi Based on Dust Events and Multi-source Remote Sensing. Acta Interdisciplinary Science. https://doi.org/10.48014/ais.20260706001
Vancouver
ZHANG Z, SONG D, SUMIYA E, WANG B, YONG M, BAO Y, et al. Fine-grained Surface Classification and Long-term Mapping of Dust Source Areas in Sino-Mongolian Gobi Based on Dust Events and Multi-source Remote Sensing. Acta Interdisciplinary Science. 2026. doi:10.48014/ais.20260706001
BibTeX
@article{ZHANG2026,
author = {ZHANG, Zirui and SONG, Dongmei and SUMIYA, Erdenesukh and WANG, Bin and YONG, Mei and BAO, Yuhai and XU, Ronghan},
title = {Fine-grained Surface Classification and Long-term Mapping of Dust Source Areas in Sino-Mongolian Gobi Based on Dust Events and Multi-source Remote Sensing},
journal = {Acta Interdisciplinary Science},
year = {2026},
doi = {10.48014/ais.20260706001}
}
EUVIMED