Mapping tropical forest biomass with radar and spaceborne LiDAR in Lopé National Park, Gabon: overcoming problems of high biomass and persistent cloud

Abstract

Abstract. Spatially-explicit maps of aboveground biomass are essential for calculating the losses and gains in forest carbon at a regional to national level. The production of such maps across wide areas will become increasingly necessary as international efforts to protect primary forests, such as the REDD+ (Reducing Emissions from Deforestation and forest Degradation) mechanism, come into effect, alongside their use for management and research more generally. However, mapping biomass over high-biomass tropical forest is challenging as (1) direct regressions with optical and radar data saturate, (2) much of the tropics is persistently cloudcovered, reducing the availability of optical data, (3) many regions include steep topography, making the use of radar data complex, (5) while LiDAR data does not suffer from saturation, expensive aircraft-derived data are necessary for complete coverage.We present a solution to the problems, using a combination of terrain-corrected L-band radar data (ALOS PAL-SAR), spaceborne LiDAR data (ICESat GLAS) and groundbased data. We map Gabon's Lopé National Park (5000 km 2 ) because it includes a range of vegetation types from savanna to closed-canopy tropical forest, is topographically complex, has no recent contiguous cloud-free high-resolution optical data, and the dense forest is above the saturation point for radar. Our 100 m resolution biomass map is derived from fusing spaceborne LiDAR (7142 ICESat GLAS footprints), 96 ground-based plots (average size 0.8 ha) and an unsupervised classification of terrain-corrected ALOS PALSAR radar data, from which we derive the aboveground biomass stocks of the park to be 78 Tg C (173 Mg C ha −1 ). This value is consistent with our field data average of 181 Mg C ha −1 , from the field plots measured in 2009 covering a total of 78 ha, and which are independent as they were not used for the GLASbiomass estimation. We estimate an uncertainty of ±25 % on our carbon stock value for the park. This error term includes uncertainties resulting from the use of a generic tropical allometric equation, the use of GLAS data to estimate Lorey's height, and the necessity of separating the landscape into distinct classes.As there is currently no spaceborne LiDAR satellite in operation (GLAS data is available for 2003-2009 only), this methodology is not suitable for change-detection. This research underlines the need for new satellite LiDAR data to provide the potential for biomass-change estimates, although this need will not be met before 2015.

Current systematic carbon-cycle observations and the need for implementing a policy-relevant carbon observing system

“…This saturation level depends on the frequency, the polarization mode, incidence angle, the type of forest, foliage structure, and moisture conditions. As a result, a wide range of sensitivities has been reported, but rarely does the sensitivity exceed 100 Mg ha −1 for L-band polarimetric algorithms (Kasischke et al, 1997;Mitchard et al, 2012). The Phased Array type L-band Synthetic Aperture Radar (PALSAR) instrument on Advanced Land Observing Satellite (ALOS) builds on the Japanese Earth Resources Satellite (JERS-1) L-band Synthetic Aperture Radar (SAR) technology, provided the first systematic global observations for generating forest change and derived biomass maps, but failed in 2011.…”

Annual South American forest loss estimates based on passive microwave remote sensing (1990–2010)

“…Other vegetation data sets that can capture vegetation dynamics are for example the observations based on long-wavelength radar backscatter (Joshi et al, 2015), where deforestation, forest degradation and the follow-up vegetation cover could be captured, and those based on observations from the SeaWinds Ku-band scatterometer (Frolking et al, 2012), which have been shown to capture gross forest loss in the Tropics. Also lidar data can be used to estimate forest biomass, and can thus capture vegetation dynamics (Mitchard et al, 2012). Data availability for radar and lidar data sets is usually from 1998 onwards.…”

Fire management in a changing landscape: a case study from Lopé National Park, Gabon

“…More accurate measures of change are required, including quantitative measures of above ground biomass and stem density, which will allow preand post-treatment comparisons. Several studies have used large-scale methods to establish landscape level biomass: measures from forest plots in Lopé have already been used to quantify satellite imagery for estimating carbon stocks at a landscape level (Mitchard et al, 2011), and this approach could be extended to improve resolution for mapping above ground biomass…”