Estimating high-density aboveground biomass within a complex tropical grassland using Worldview-3 imagery

Area of Science:

• Ecology
• Remote Sensing
• Agricultural Science

Background:

• Grassland ecosystems face degradation due to urbanization and agriculture.
• Nitrogen fertilization is used for grassland rehabilitation, necessitating effective monitoring.
• Biomass signal saturation in traditional indices hinders monitoring in dense vegetation.

Purpose of the Study:

• To evaluate Worldview-3 imagery for monitoring high-density tropical grasslands.
• To reduce biomass signal saturation using modified vegetation indices and image processing.
• To develop a cost-effective framework for rangeland management.

Main Methods:

• Utilized high-resolution Worldview-3 imagery.
• Developed modified NDVI-type indices using random forest algorithm.
• Applied linear contrast stretching, histogram equalization, and Singular Value Decomposition (SVD).

Main Results:

• Contrast enhancement techniques combined with SVD improved biomass estimation.
• Modified NDVI indices from red, green, and NIR1 bands showed best results (R²=0.71).
• The developed framework effectively addressed biomass saturation issues.

Conclusions:

• Worldview-3 imagery and modified indices offer a solution to biomass saturation.
• The framework aids rangeland managers in assessing vegetation condition in high-density grasslands.
• This approach provides a valuable tool for sustainable rangeland management.