FEBRUARY 18, 2000

Canopy Height Models and Airborne Lasers to Estimate Forest
Biomass

Airborne lasers can be used to measure tree heights, height variability,
and tree canopy closure over extensive areas (100's or 1000's of sq. km)
rapidly. An airborne laser flown over a forest collects ranging data from
the aircraft to the top of the canopy and from the aircraft to ground.
These forest height measurements describe a transect, i.e., a linear sample,
through the forest, and this transect sample of height measurements can be
used to calculate forest biomass for the region.

In order to estimate forest biomass using airborne laser ranging
measurements, an equation which predicts biomass as a function of laser
measurements is needed. Such an equation can be calculated by 1) taking
forest measurements on numerous ground plots, 2) estimating the biomass on
those plots, 3) deducing what a laser would have measured from the air over
those same ground plots, and 4) predicting biomass as a function of laser
variables. Step 3 can be accomplished using canopy simulation techniques
whereby the ground data are used to reconstruct the forest canopy in a
computer.

A study in Costa Rica concluded that this canopy simulation approach to
estimating forest biomass using airborne lidar was often very inaccurate.
An error analysis undertaken to identify the sources of the inaccuracies
found that the width of the fixed-area plot used to collect the tree data on
the ground was critical. Plot widths less than 6-8m produced forest canopy
simulations that were more open (less dense) and shorter than the actual
forest stands. The inaccurate recreation of the forest canopy propagated
through the rest of the procedure, leading to inaccurate laser estimates of
forest biomass.

If the appropriate ground sampling procedures are utilized, then results
suggest that airborne laser altimetry will produce biomass estimates within
20-25% of the comparable ground values in primary tropical forest. Airborne
lasers should be considered for use as a reconnaissance tool in forested
regions where little or no biomass information exists.

Reference: Nelson, R., J. Jimenez, C. Schnell, G.S. Hartshorn, T.G.
Gregoire, and R. Oderwald. Canopy Height Models and Airborne Lasers to
Estimate Forest Biomass: 2 Problems. International Journal of Remote
Sensing, in press.