Postfire stand structure in a semiarid savanna: cross-scale challenges estimating biomass

Area of Science:

• Ecology
• Remote Sensing
• Forestry

Background:

• Large-scale carbon stock inventories rely on algorithms linking remotely sensed woody cover to biomass.
• These algorithms are often applied generically, ignoring disturbances like fire that alter vegetation structure.
• Disturbances can significantly impact the relationship between woody cover and biomass.

Purpose of the Study:

• To compare field and remote sensing estimates of woody biomass on savannas with different fire histories.
• To assess potential errors in estimating woody biomass from cover without considering fire history.
• To evaluate the accuracy of generic two-dimensional (2-D) cover algorithms versus disturbance-specific models.

Main Methods:

• Field surveys quantified multilayer cover (MLC) on burned and unburned savanna sites.
• Remote sensing derived woody cover fraction (WCF) using Landsat Thematic Mapper imagery and spectral mixture analysis.
• Allometric relationships were used to relate field and satellite estimates to aboveground biomass.

Main Results:

• Woody cover was similar on sites burned 11-16 years prior and unburned control sites.
• Aboveground biomass was approximately twofold higher on unburned control sites compared to burned sites.
• Fire history altered the cover-biomass relationship from linear (control) to curvilinear (burned).

Conclusions:

• Generic 2-D cover algorithms may underestimate biomass in undisturbed stands and overestimate it in recovering stands.
• Accurate woody biomass estimation requires accounting for fire history and potentially vegetation height.
• Disturbance-specific models or 3-D vegetation volume estimations are needed for improved accuracy.