An analysis of forest biomass sampling strategies across scales

Abstract

Abstract. Tropical forests play an important role in the global carbon cycle as they store a large amount of carbon in their biomass. To estimate the mean biomass of a forested landscape, sample plots are often used, assuming that the biomass of these plots represents the biomass of the surrounding forest. In this study, we investigated the conditions under which a limited number of sample plots conform to this assumption. Therefore, the minimum number of sample sizes for predicting the mean biomass of tropical forest landscapes was determined by combining statistical methods with simulations of sampling strategies. We examined forest biomass maps of Barro Colorado Island (50 ha), Panama (50 000 km2), and South America, Africa, and Southeast Asia (3 × 106–11 × 106 km2). The results showed that around 100 plots (1–25 ha each) are necessary for continent-wide biomass estimations if the sampled plots are randomly distributed. However, locations of current inventory plots often do not meet this requirement, for example, as their sampling design is based on spatial transects among climatic gradients. We show that these nonrandom locations lead to a much higher sampling intensity being required (up to 54 000 plots for accurate biomass estimates for South America). The number of sample plots needed can be reduced using large distances (5 km) between the plots within transects. We also applied novel point pattern reconstruction methods to account for aggregation of inventory plots in known forest plot networks. The results implied that current plot networks can have clustered structures that reduce the accuracy of large-scale estimates of forest biomass if no further statistical approach is applied. To establish more reliable biomass predictions across South American tropical forests, we recommend more spatially randomly distributed inventory plots (minimum: 100 plots) and ensuring that the analyses of inventory plot data consider their spatial characteristics. The precision of forest attribute estimates depends on the sampling intensity and strategy.

Natural forests in New Zealand – a large terrestrial carbon pool in a national state of equilibrium

“…Moreover, the inclusion of the 20 m radius plot data in the analysis avoids a bias in large tree stem numbers associated with using only the 20 m × 20 m square plots (Paul et al 2019) which is the reason it results in a reduction in our estimate of total carbon. Our use of larger plots also aligns us with research that identified smaller plot size as an issue when monitoring forest dynamics (Clark and Clark 2000;Wagner et al 2010;McMahon et al 2019;Hetzer et al 2020). Inclusion of these additional data also provided more precise estimates of carbon stocks and stock changes.…”

Mapping Amazon Forest Productivity by Fusing GEDI Lidar Waveforms with an Individual-Based Forest Model

“…However, a wider range of GPP values was observed in this study (0-40 Mg C ha −1 a −1 ) in comparison to MODIS product (values between 15 and 35). The data of Malhi et al [32] showed higher values than the average GPP determined in this study, but this was reasonable because their amount of included forest inventory data was rather low, the forest plots were much smaller in size than the resolution of the our map (1 km 2 ), and the locations of these forest plots were often in productive forests and not randomly selected [36]. , the MODIS-derived GPP product, and forest inventory data from Malhi et al [32] for the Amazon Basin.…”

Assessing scale‐dependent effects on Forest biomass productivity based on machine learning

“…Indeed, determining the optimum sampling area depends on the manpower and material resources consumed by the field survey and the accuracy required by the investigators. Hetzer et al (2020) showed that 1 ha is the effective area for mean biomass estimation with sufficient precision in South America (Hetzer et al, 2020). We consider a sampling area of at least 140 × 140 m to reach the requirements for estimates of biomass productivity in this region (Figure 4 and Figure ), as the turning point in R 2 and rRMSE indicates the minimal scale for effective sampling is the above scale, sampling at a median scale may be more cost effective ( R 2 > 0.7, rRMSE <20%).…”

“…The spatial heterogeneity of sampling units can cause difficulties for community surveys, but large sampling areas are fundamental for reflecting community characteristics. Indeed, determining the optimum sampling area depends on the manpower and material resources consumed by the et al (2020) showed that 1 ha is the effective area for mean biomass estimation with sufficient precision in South America (Hetzer et al, 2020). We consider a sampling area of at least 140 × 140 m to reach the requirements for estimates of biomass productivity in this region (Figure 4 previous research may have some problems.…”