1. Swedish University of Agricultural Sciences www.slu.se Uncertainty and verification of soil organic carbon change estimates Erik Karltun and Carina Ortiz Swedish University of Agricultural Sciences

2. Swedish University of Agricultural Sciences www.slu.se Carbon in Swedish soils -The carbon stock of Swedish forest soils estimated at 2000 Mton carbon (DOM and SOC) -An annual change of 1 equates to 2 Mton C or 7 Mton CO 2, i.e. > 10% of Swedish national GHG emissions -Soils are both sinks and sources -Reporting needs monitoring and verification – do we have appropriate tools?

3. Swedish University of Agricultural Sciences www.slu.se Integrated with the NFI Total number of sampled plots c.4500 DOM and SOC sampled separately Soil type according to World Reference Base Organic soil=Histosols Forest Soil Inventory

4. Swedish University of Agricultural Sciences www.slu.se B hor sampled BC sampled 50 cm 0 cm DOM C sampled Interpol. E hor.

5. Swedish University of Agricultural Sciences www.slu.se 9091939492959697049900010298030506070809 2 nd Soil Survey3 rd Soil SurveyNo data Extrapol. Sampled UNFCCC-Kyoto reporting ResampledInterpolated Data Reporting timeline

6. Swedish University of Agricultural Sciences www.slu.se Two views on soil carbon change

7. Swedish University of Agricultural Sciences www.slu.se Models – an alternative way -Carbon pools and carbon pool changes was estimated using to conceptually different dynamic carbon models, Q and Yasso07 -Carbon pool development simulated from 1926 using climate and litter input data -Uncertainty of model output determined by parameter uncertainty analysis and Monte Carlo simulation methodology

8. Swedish University of Agricultural Sciences www.slu.se 500 1000 1500 2000 2500 3000 3500 1920193019401950196019701980199020002010 Stående skogsvolym – m 3 Källa: Riksskogstaxeringen, SLU Standing forest volume

9. Swedish University of Agricultural Sciences www.slu.se

10. Swedish University of Agricultural Sciences www.slu.se

11. Swedish University of Agricultural Sciences www.slu.se Conclusions -Small changes in the soil pool can have a considerable impact on the national GHG balance -Over a timeperiod of more than 10 years both models and measurement indicate that mineral soil is a sink but uncertainty is high -The possibility to monitor and verify small changes is limited due to method uncertainty -Soils seem to respond slowly to increase in forest biomass -Combination of various approaches (measurements, models valuable) -Uncertainty reporting must be improved – todays reporting of uncertainty is not relevant to LULUCF