[Land salinization information extraction method based on HSI hyperspectral and TM imagery].

Area of Science:

• Environmental Remote Sensing
• Geospatial Analysis
• Soil Science

Context:

• Coastal salinization poses a significant threat to agricultural productivity and ecosystem health in the Yellow River Delta.
• Accurate mapping of salinized land is crucial for effective land management and resource utilization.

Purpose:

• To evaluate the efficacy of hyperspectral imagery (HSI) versus multispectral imagery (MSI) for extracting land salinization information.
• To develop and validate a land salinization information extraction model using spectral characteristics and soil salinity data.

Summary:

• This study utilized HJ-1A HSI and Landsat TM MSI data, pre-processed and analyzed for spectral characteristics of various land use types, including different degrees of salinization.
• A decision tree classification method, integrating spectral data and soil salinity content, was employed for land salinization extraction.
• Results indicated higher overall accuracy (96.43%) and Kappa coefficient (95.59%) for HSI compared to MSI (89.17% and 86.74%), demonstrating HSI's greater precision in distinguishing soil salinity levels.

Impact:

• Hyperspectral imagery provides a more accurate and efficient method for extracting land salinization information than multispectral imagery.
• The study provides a valuable decision-making basis for the scientific utilization and management of coastal salinized land resources in the Yellow River Delta.