Sources of errors and uncertainties in the assessment of forest soil carbon stocks at different scales—review and recommendations

Abstract

Spatially explicit knowledge of recent and past soil organic carbon (SOC) stocks in forests will improve our understanding of the effect of human- and non-human-induced changes on forest C fluxes. For SOC accounting, a minimum detectable difference must be defined in order to adequately determine temporal changes and spatial differences in SOC. This requires sufficiently detailed data to predict SOC stocks at appropriate scales within the required accuracy so that only significant changes are accounted for. When designing sampling campaigns, taking into account factors influencing SOC spatial and temporal distribution (such as soil type, topography, climate and vegetation) are needed to optimise sampling depths and numbers of samples, thereby ensuring that samples accurately reflect the distribution of SOC at a site. Furthermore, the appropriate scales related to the research question need to be defined: profile, plot, forests, catchment, national or wider. Scaling up SOC stocks from point sample to landscape unit is challenging, and thus requires reliable baseline data. Knowledge of the associated uncertainties related to SOC measures at each particular scale and how to reduce them is crucial for assessing SOC stocks with the highest possible accuracy at each scale. This review identifies where potential sources of errors and uncertainties related to forest SOC stock estimation occur at five different scales-sample, profile, plot, landscape/regional and European. Recommendations are also provided on how to reduce forest SOC uncertainties and increase efficiency of SOC assessment at each scale.

The devil is in the detail: Discrepancy between soil organic carbon stocks estimated from regional and local data sources in Flanders, Belgium

“…Furthermore, quantitative estimates of the various sources of uncertainty present at different scales (see e.g., Goidts, van Wesemael, and Crucifix () and Vanguelova et al. ()) would serve in‐depth comparison between assessments. Several methods exist to estimate uncertainty, for example through stochastic Monte‐Carlo simulations (Chartin et al., ) or decision trees (Stumpf et al., ).…”

Concept and Methodology of the National Forest Soil Inventory

“…The derivation of trends in element stocks from element concentration and fine earth content measurements at two points in time is challenging due to the numerous possibilities for measurement errors. High spatial and temporal variability of element concentrations on the same plot, the representativeness of subsamples of sieved soil taken for measurements, the relevance of extremely low concentrations in the deeper soil layers, the difficulty of accurate volumetric estimations of fine earth contents and bulk density of soil samples, the variable accuracy of laboratory measurements at both points in time and the inaccuracy of depth delineations are the main error sources (Vanguelova et al 2016).…”

Spatial Uncertainty Propagation Analysis with the spup R Package

“…Hengl et al, 2010;Hamel and Guswa, 2015;Muthusamy et al, 2016;Nijhof et al, 2016;Yu et al, 2016), biogeochemistry (e.g. Boyer et al, 2006;Nol et al, 2010;Pennock et al, 2010;Hugelius, 2012;Vanguelova et al, 2016), or ecology and conservation planning (e.g. Jager et al, 2005;Fisher et al, 2010;Lechner et al, 2014).…”