136-3 Mapping Organic Carbon Stocks of Swiss Forest Soils.

See more from this Division: S05 Pedology
See more from this Session: Symposium--Global Soil Mapping in a Changing World
Monday, October 22, 2012: 3:10 PM
Duke Energy Convention Center, Room 264, Level 2
Share |

Madlene Nussbaum1, Andreas Papritz1, Andri Baltensweiler2 and Lorenz Walthert3, (1)Institute of Terrestrial Ecosystems ITES, Swiss Federal Institute of Technology ETH, Zurich, Switzerland
(2)Forest Resources and Management, Swiss Federal Institute for Forest, Snow and Landscape Research (WSL), Birmensdorf, Switzerland
(3)Forest Soils and Biogeochemistry, Swiss Federal Institute for Forest, Snow and Landscape Research (WSL), Birmensdorf, Switzerland
Carbon (C) sequestration in forests offsets greenhouse gas emissions. Therefore, quantifying C stocks and fluxes in forest ecosystems is of interest for greenhouse gas reporting according to the Kyoto protocol. In Switzerland, the National Forest Inventory offers comprehensive data to quantify the aboveground forest biomass and its change in time. Estimating stocks of soil organic C (SOC) in forests is more difficult because of its high spatial variability and the relatively imprecise determinability of some of the involved soil properties.

Based on data from 1’033 plots we modeled SOC stocks of the organic layer and the mineral soil to depths of 30 cm and 100 cm for the Swiss forested area. We applied a novel robust restricted maximum likelihood method to fit a linear regression model with spatially correlated errors. For the regression analysis we used a broad range of covariates derived from climate data (e. g. precipitation, temperature), two elevation models (resolutions 25 and 2 m) with their terrain attributes and spectral reflectance data representing vegetation. Furthermore, the main cartographic categories of an overview soil map and a small-scale geological map were used to coarsely represent the parent material.

Results for topsoil SOC (0-30 cm without organic surface layer) showed residual autocorrelation that was weak but significant. Precipitation, spectral reflectance of the vegetation in near-infrared wavelength, a topographic position index and aggregated soil and geological map information were the only significant covariates. Testing the predictive power of the fitted model with independent test data resulted in satisfactory precision of the predictions (coefficient of determination 0.34). The fitted model was used to compute a robust kriging prediction map of the carbon stock in forest topsoils on a 1-ha grid over Switzerland. The mean predicted SOC stock to a depth of 30 cm amounts to 79.9 Mg/ha (coefficient of variation 0.34).

See more from this Division: S05 Pedology
See more from this Session: Symposium--Global Soil Mapping in a Changing World