Combining remote sensing imagery and forest age inventory for biomass mapping

Area of Science:

• Forestry
• Remote Sensing
• Ecology
• Carbon Cycle Research

Background:

• Aboveground biomass (AGB) is crucial for understanding the global carbon cycle.
• Accurate AGB estimation is vital for forest management and climate change studies.
• Remote sensing offers a scalable approach for AGB assessment.

Purpose of the Study:

• To estimate forest aboveground biomass (AGB) in Liping County, Guizhou Province, China.
• To develop and validate AGB estimation models using Landsat ETM(+) imagery and field data.
• To analyze the influence of forest type on AGB prediction accuracy.

Main Methods:

• Calculation of vegetation indices (SR, RSR, NDVI) from atmospherically corrected satellite imagery.
• Development of a Leaf Area Index (LAI) map using regression analysis.
• Integration of vegetation indices, LAI, and forest stand age into stepwise regression models for AGB estimation.
• Stratification of models by forest type (Chinese fir, coniferous, broadleaved, mixed).

Main Results:

• Models incorporating LAI and NDVI achieved high accuracy for Chinese fir (R²=0.93).
• Coniferous forests showed 94% AGB variance explained by LAI and stand age.
• Broadleaved forests' AGB variance was primarily explained by stand age (R²=0.792).
• Mixed forests' AGB was well predicted by LAI and SR (R²=0.931).
• A general model using SR and LAI explained 90% of AGB variance across all forest types.
• A distinct northeast-to-southwest gradient in AGB was observed.

Conclusions:

• The study successfully demonstrated the utility of remote sensing and field data for accurate forest AGB estimation.
• Predictive power of variables for AGB varies significantly across different forest types.
• Developed models provide valuable tools for monitoring forest carbon stocks in the region.