3.3.4 Biomass from Lidar
Recently lidar data have become widely available to study the linkage between
lidar signals and vegetation structure characteristics. AGB is strongly related to
lidar measured tree height, ranging from boreal conifers to equatorial rain forests.
Lidar data are mainly from airborne discrete-return lidar (Lim and Treitz 2004;
Lim et al. 2003; Popescu 2007; García et al. 2010), airborne profiling lidar (e.g.,
Nelson et al. 1988), airborne waveform lidar (Drake et al. 2002; Lefsky et al.
1999a, b, c; Ni-Meister et al. 2010), satellite waveform lidar (Boudreau et al. 2008;
Nelson 2010; Lefsky et al. 2002, 2005), and ground-based lidar (Ni-Meister et al.
2010). Lidar is recognized as the state-of-the-art remote sensing technology for
mapping AGB because it is much less sensitive to the saturation problem, compared to conventional optical remote sensing and radar data. In the following, we
summarize recent progress on lidar-based biomass mapping activities from small
footprint discrete return lidar and large footprint full waveform lidar.
3.3.4.1 Small Footprint Discrete Return Lidar
Small-footprint discrete multiple return lidar data has been collected in many small
regions of the globe. Such small footprint airborne lidar systems are available on a
commercial basis and are now operationally used for forest resource inventories
(Næsset and Gobakken 2008). With many ground lidar systems, complex and
detailed vegetation structure data have been recorded over various study sites.
These global, regional, and local lidar data can provide the detailed vegetation
structure and biomass maps necessary for carbon models and ecosystem processes
studies.
AGB has been successfully estimated using small footprint discrete lidar data
(Lim et al. 2003; Næsset 2004; Nelson et al. 1988; Popescu et al. 2007; García
et al. 2010). Tree height obtained from airborne lidar is a good predictor of
biomass for large area averages (Nelson et al. 2003, 2004), which can explain 88
and 85 % of the variability in aboveground and belowground biomass, respectively, for 1,395 sample plots in the coniferous boreal zone of Norway (Næsset and
Gobakken 2008). Regardless of the type of lidar system used, however, estimation
of biomass is generally conducted based on regression equations relating vegetation biomass to lidar-derived variables across different scales from individual
tree to plot and stand scales.
At plot scale, field-measured biomass is regressed against derived statistics
from plot-level lidar data. The lidar statistics can be obtained from the individual
returns or from the height of canopy (also called canopy height model (CHM))
where lidar return values are interpolated to a certain size raster. This approach
adopts distributional metrics such as the mean canopy height and the standard
deviation of the canopy height derived from CHM or the raw returns. These
metrics are then used in conjunction with regression equations to predict forest
3 Remote Sensing of Forest Biomass
81
Recently lidar data have become widely available to study the linkage between
lidar signals and vegetation structure characteristics. AGB is strongly related to
lidar measured tree height, ranging from boreal conifers to equatorial rain forests.
Lidar data are mainly from airborne discrete-return lidar (Lim and Treitz 2004;
Lim et al. 2003; Popescu 2007; García et al. 2010), airborne profiling lidar (e.g.,
Nelson et al. 1988), airborne waveform lidar (Drake et al. 2002; Lefsky et al.
1999a, b, c; Ni-Meister et al. 2010), satellite waveform lidar (Boudreau et al. 2008;
Nelson 2010; Lefsky et al. 2002, 2005), and ground-based lidar (Ni-Meister et al.
2010). Lidar is recognized as the state-of-the-art remote sensing technology for
mapping AGB because it is much less sensitive to the saturation problem, compared to conventional optical remote sensing and radar data. In the following, we
summarize recent progress on lidar-based biomass mapping activities from small
footprint discrete return lidar and large footprint full waveform lidar.
3.3.4.1 Small Footprint Discrete Return Lidar
Small-footprint discrete multiple return lidar data has been collected in many small
regions of the globe. Such small footprint airborne lidar systems are available on a
commercial basis and are now operationally used for forest resource inventories
(Næsset and Gobakken 2008). With many ground lidar systems, complex and
detailed vegetation structure data have been recorded over various study sites.
These global, regional, and local lidar data can provide the detailed vegetation
structure and biomass maps necessary for carbon models and ecosystem processes
studies.
AGB has been successfully estimated using small footprint discrete lidar data
(Lim et al. 2003; Næsset 2004; Nelson et al. 1988; Popescu et al. 2007; García
et al. 2010). Tree height obtained from airborne lidar is a good predictor of
biomass for large area averages (Nelson et al. 2003, 2004), which can explain 88
and 85 % of the variability in aboveground and belowground biomass, respectively, for 1,395 sample plots in the coniferous boreal zone of Norway (Næsset and
Gobakken 2008). Regardless of the type of lidar system used, however, estimation
of biomass is generally conducted based on regression equations relating vegetation biomass to lidar-derived variables across different scales from individual
tree to plot and stand scales.
At plot scale, field-measured biomass is regressed against derived statistics
from plot-level lidar data. The lidar statistics can be obtained from the individual
returns or from the height of canopy (also called canopy height model (CHM))
where lidar return values are interpolated to a certain size raster. This approach
adopts distributional metrics such as the mean canopy height and the standard
deviation of the canopy height derived from CHM or the raw returns. These
metrics are then used in conjunction with regression equations to predict forest
3 Remote Sensing of Forest Biomass
81