Detection of large above-ground biomass variability in lowland forest ecosystems by airborne LiDAR

Abstract

Quantification of tropical forest above-ground biomass (AGB) over large areas as input for Reduced Emissions from Deforestation and forest Degradation (REDD+) projects and climate change models is challenging. This is the first study which attempts to estimate AGB and its variability across large areas of tropical lowland forests in Central Kalimantan (Indonesia) through correlating airborne light detection and ranging (LiDAR) to forest inventory data. Two LiDAR height metrics were analysed, and regression models could be improved through the use of LiDAR point densities as input (<i>R</i>² = 0.88; <i>n</i> = 52). Surveying with a LiDAR point density per square metre of about 4 resulted in the best cost / benefit ratio. We estimated AGB for 600 km of LiDAR tracks and showed that there exists a considerable variability of up to 140% within the same forest type due to varying environmental conditions. Impact from logging operations and the associated AGB losses dating back more than 10 yr could be assessed by LiDAR but not by multispectral satellite imagery. Comparison with a Landsat classification for a 1 million ha study area where AGB values were based on site-specific field inventory data, regional literature estimates, and default values by the Intergovernmental Panel on Climate Change (IPCC) showed an overestimation of 43%, 102%, and 137%, respectively. The results show that AGB overestimation may lead to wrong greenhouse gas (GHG) emission estimates due to deforestation in climate models. For REDD+ projects this leads to inaccurate carbon stock estimates and consequently to significantly wrong REDD+ based compensation payments

Current remote sensing approaches to monitoring forest degradation in support of countries measurement, reporting and verification (MRV) systems for REDD+

“…), r 2 0.72 (early)RH50 r 2 0.3 (intermed. )[11] Canopy height changeRBrazil1[2]RGabonr 2  = 0.83, RMSE 3.3 m, n = 95[65]RAustralia[55]RCameroon, DRC275–82%[13]RTanzania[84]NUganda0.9 mm (bias), 8–16 mm (bias upslope)[85] AGB changeRKalimantanr 2  = 0.77 (PPR 54.2 Mg ha −1 ), r 2  = 0.81 (PPR 47.4 Mg ha −1 )[16]RBrazilR 2  = 0.7, SE 41.5 Mg ha −1 [2]RKalimantanr 2  = 0.88, RMSE ± 13.8 Mg 0.13 ha −1 [46]RNorway1495.7–97.8% (classification of deforestation and untouched classes), 56.3–69.2% (degradation classes)AGB: SE 5–8.4 Mg ha −1 , r 2  = 0.88–0.98SE reduced by 18–84% using LiDAR, largest gains in degradation class (73–84%)[67]RKalimantan50/100SE 53.2 Mg ha −1 (n = 51 @50 m), 49.1 Mg ha −1 (@100 m)[71]RBrazil30p < 0.0001, R 2  = 0.6, N = 26[29]R…”

“…The change in AGB associated with the disturbed forest area was estimated at −17.9 ± 3.1 Mg ha −1 (p < 0.0001). Jubanski et al [46] assessed the variability in AGB in lowland tropical forests in Kalimantan. LiDAR-derived height metrics correlated well with model-based estimates of AGB (r 2 of 0.88, RMSE ± 13.79 Mg 0.13 ha −1 ).…”

Estimating Logged-Over Lowland Rainforest Aboveground Biomass in Sabah, Malaysia Using Airborne LiDAR Data

“…Mean height metric from LiDAR data has widely been used to develop the AGB or carbon estimation models: lowland rainforest in Central Kalimantan (Jubanski et al 2013), rainforest in Panama, Peru, Madagascar, and Hawaii (Asner et al 2012c), and tropical montane forest in Sabah, Borneo . Although the model's coefficient of determination was not very high (R 2 = 0.67), the single-variable model had an RMSE of 22.31% of the mean AGB, which is lower than the RMSE of the single-variable model using mean canopy height (28% of the mean AGB) and multiplevariable model (26% of the mean AGB) .…”

“…Multiple linear regression analysis was then performed to examine any further model improvement by incorporating multiple LiDAR variables. The simple regression analyses were carried out with power models because the power models were successfully used to estimate AGB in tropical forests (Asner et al 2012a, b;Jubanski et al 2013). In the multiple regression analysis all LiDAR variables were transformed using the natural logarithm and stepwise regression using the Akaike Information Criterion (AIC) conducted to determine the final model.…”

“…Asner et al (2012c) examined the effectiveness of the extracted mean canopy height to estimate AGB in Panama, Peru, Madagascar, and Hawaii (R 2 = 0.8, RMSE = 27.6 Mg C ha -1 ). In Borneo, Kronseder et al (2012) and Jubanski et al (2013) examined the use of LiDAR derived height metrics to estimate the AGB in logged peat swamp forest and unlogged lowland dipterocarp forest in Central Kalimantan. Ioki et al (2014) tested the use of laser penetration rate from LiDAR data for AGB estimation of primary and degraded tropical montane forests of in Sabah, Borneo (R 2 = 0.78, RMSE = 27.6 t ha -1 ).…”