Scholarlink

<p>Accurate information on soil properties throughout the soil profile is essential for effective soil and water management in dryland regions. This study aimed to digitally map key soil physical and chemical properties important for plant growth and water retention in dry environments from the surface to 1.5&#xa0;m depth in the Central Dry Zone of Myanmar and to evaluate the contribution of Landsat spectral data to model performance across different soil depths. A total of 4,059 soil samples from 800 sites were collected across seven depth intervals (0–150 cm). Soil properties, including texture, organic carbon, pH, electrical conductivity, effective cation exchange capacity, exchangeable cations, exchangeable sodium percentage, rock fragment content, were modelled using Random Forest algorithms, and soil depth was modelled separately using a Cubist model. Landsat spectral bands and terrain attributes derived from a digital elevation model were used as covariates. Landsat data capture vegetation, moisture, and surface reflectance, representing organisms, and soil condition, while terrain attributes represent relief-driven processes such as water redistribution, erosion, and deposition. Profile available water capacity was estimated using pedotransfer functions based on predicted soil properties. Soil properties showed strong vertical organisation, with organic carbon and sand concentrated in surface layers and increasing clay content with depth. Incorporating Landsat spectral data significantly improved prediction accuracy for surface and upper subsoil layers (0–30 cm), particularly for texture, organic carbon, and effective cation exchange capacity. Model improvements declined progressively with depth. Spatial predictions revealed coherent landscape-scale patterns in soil properties, soil depth, rock fragments, and available water capacity. The results demonstrate that optical remote sensing enhances digital soil mapping in surface and upper subsoil layers but has limited value at greater depths. The study provides the first region-wide depth-explicit digital maps of soil properties and available water capacity maps for the Central Dry Zone of Myanmar and offers a transferable framework for soil assessment in data-scarce dryland regions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Digital Mapping of Soil Profile Properties using Remote Sensing and Topography Covariates in the Central Dry Zone of Myanmar

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月14日

查看全文
  • Akari Win,
  • Budiman Minasny,
  • Anthony Ringrose-Voase,
  • Ho Jun Jang

摘要

Accurate information on soil properties throughout the soil profile is essential for effective soil and water management in dryland regions. This study aimed to digitally map key soil physical and chemical properties important for plant growth and water retention in dry environments from the surface to 1.5 m depth in the Central Dry Zone of Myanmar and to evaluate the contribution of Landsat spectral data to model performance across different soil depths. A total of 4,059 soil samples from 800 sites were collected across seven depth intervals (0–150 cm). Soil properties, including texture, organic carbon, pH, electrical conductivity, effective cation exchange capacity, exchangeable cations, exchangeable sodium percentage, rock fragment content, were modelled using Random Forest algorithms, and soil depth was modelled separately using a Cubist model. Landsat spectral bands and terrain attributes derived from a digital elevation model were used as covariates. Landsat data capture vegetation, moisture, and surface reflectance, representing organisms, and soil condition, while terrain attributes represent relief-driven processes such as water redistribution, erosion, and deposition. Profile available water capacity was estimated using pedotransfer functions based on predicted soil properties. Soil properties showed strong vertical organisation, with organic carbon and sand concentrated in surface layers and increasing clay content with depth. Incorporating Landsat spectral data significantly improved prediction accuracy for surface and upper subsoil layers (0–30 cm), particularly for texture, organic carbon, and effective cation exchange capacity. Model improvements declined progressively with depth. Spatial predictions revealed coherent landscape-scale patterns in soil properties, soil depth, rock fragments, and available water capacity. The results demonstrate that optical remote sensing enhances digital soil mapping in surface and upper subsoil layers but has limited value at greater depths. The study provides the first region-wide depth-explicit digital maps of soil properties and available water capacity maps for the Central Dry Zone of Myanmar and offers a transferable framework for soil assessment in data-scarce dryland regions.