Asosiy kontentga oʻtish
Maqola

COMPARATIVE SOIL EROSION SUSCEPTIBILITY ASSESSMENT USING STATISTICAL AND DATA-DRIVEN TECHNIQUES: EVIDENCE FROM THE CHARTAK DISTRICT, UZBEKISTAN

M. KHOLMURODOVAInstitute of Fundamental and Applied Research , "TIIAME" National Research University , Kori Niyoziy 39A , Tashkent 100000 , UzbekistanM. JULIEVSH. KHAMIDULLAEVZ. DJANPULATOVAInstitute of Fundamental and Applied Research , "TIIAME" National Research University , Kori Niyoziy 39A , Tashkent 100000 , UzbekistanS. TURDALIEVAInstitute of Fundamental and Applied Research , "TIIAME" National Research University , Kori Niyoziy 39A , Tashkent 100000 , UzbekistanZ. KHADJIEVAInstitute of Fundamental and Applied Research , "TIIAME" National Research University , Kori Niyoziy 39A , Tashkent 100000 , UzbekistanB. ABDIKAIROVInstitute of Agriculture and Agrotechnologies of Karakalpakstan , Abdametov Street , Nukus 230109 , UzbekistanM. SHAMSIEVO. KHAKBERDIEVNational University of Uzbekistan
2026
ABI

Annotatsiya

Soil erosion is a significant environmental problem that negatively affects agriculture and land resources.This study aimed to evaluate soil erosion susceptibility using Geographic Information Systems (GIS) and three bivariate statistical models: Statistical Index (SI), Frequency Ratio (FR), and Certainty Factor (CF).A total of 133 erosion points within the study area were used as reference data.Remote sensing and topographic datasets were used to derive nine widely applied environmental conditioning factors slope, aspect, plan curvature, elevation, stream power index (SPI), drainage density, distance to rivers, normalized difference vegetation index (NDVI), and land use/land cover (LULC).These factors were selected based on their frequent use and proven effectiveness in recent erosion-susceptibility studies.Raster layers with a resolution of 30 m were produced for each factor.The susceptibility maps indicated that large portions of the study area fall within the high to very high susceptibility categories.This is especially true for rangelands, steep slopes, areas with sparse vegetation.Quantitative validation using the Area Under the Receiver Operating Characteristic Curve (ROC-AUC) indicated acceptable predictive performance: 0.879 for SI, 0.843 for FR, and 0.878 for CF.The SI (0.879) and CF (0.878) models yielded the highest AUC values, whereas the FR model produced a lower value (0.843).All three models demonstrated strong predictive capability based on ROC-AUC metrics.Overall, the results underscore the importance of integrating both statistical models and GIS techniques to assess soil erosion susceptibility.The generated maps can help decision-makers develop soil conservation strategies, prioritize high-risk areas, and promote sustainable land use.

Hali tarjima qilinmagan

Identifikatorlar

Iqtiboslar va manbalar

0 ta iqtibos0 ta foydalanilgan manba