M. A. Vanyarkho, A. Sh. Ziyatdinov, I. L. Prokofiev, N. L. Fedorova, S. S. Ulanova
{"title":"Method of Determining the Degree of Fixation of Sands in the Desert Zone of Kalmykia by Vegetation Using Satellite Radar","authors":"M. A. Vanyarkho, A. Sh. Ziyatdinov, I. L. Prokofiev, N. L. Fedorova, S. S. Ulanova","doi":"10.1134/S2079096126700149","DOIUrl":null,"url":null,"abstract":"<p>A method for assessing land degradation in the desert zone of Kalmykia has been developed, combining radar and multispectral remote sensing data. The basis is the assessment of the projective cover of vegetation and the change in nanorelief using the interferometric coherence method, supplemented by spectral indices (SAVI, LAI, shortwave albedo). The resulting feature space allows us to classify a section of territory into one of five possible types of objects: salt marshes, open sands, and weakly, moderately, or densely overgrown sands. The classification accuracy of the trained neural network was confirmed by field data with a correspondence of 70 to 100% depending on the class, with the average accuracy being 90%.</p>","PeriodicalId":44316,"journal":{"name":"Arid Ecosystems","volume":"16 1","pages":"132 - 141"},"PeriodicalIF":0.6000,"publicationDate":"2026-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Arid Ecosystems","FirstCategoryId":"1085","ListUrlMain":"https://link.springer.com/article/10.1134/S2079096126700149","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"ECOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
A method for assessing land degradation in the desert zone of Kalmykia has been developed, combining radar and multispectral remote sensing data. The basis is the assessment of the projective cover of vegetation and the change in nanorelief using the interferometric coherence method, supplemented by spectral indices (SAVI, LAI, shortwave albedo). The resulting feature space allows us to classify a section of territory into one of five possible types of objects: salt marshes, open sands, and weakly, moderately, or densely overgrown sands. The classification accuracy of the trained neural network was confirmed by field data with a correspondence of 70 to 100% depending on the class, with the average accuracy being 90%.
期刊介绍：
Arid Ecosystems publishes original scientific research articles on desert and semidesert ecosystems and environment:systematic studies of arid territories: climate changes, water supply of territories, soils as ecological factors of ecosystems state and dynamics in different scales (from local to global);systematic studies of arid ecosystems: composition and structure, diversity, ecology; paleohistory; dynamics under anthropogenic and natural factors impact, including climate changes; studying of bioresources and biodiversity, and development of the mapping methods;arid ecosystems protection: development of the theory and methods of degradation prevention and monitoring; desert ecosystems rehabilitation;problems of desertification: theoretical and practical issues of modern aridization processes under anthropogenic impact and global climate changes.
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
