Regional assessments of change in soil organic carbon (SOC) stocks due to land-use change are essential for supporting policy and management decisions related to greenhouse gas emissions and mitigation through carbon sequestration in soils. We have developed an improved approach by integrating geostatistical techniques with the Intergovernmental Panel on Climate Change (IPCC) carbon inventory approach to assess the impact of no-till management and crop-residue retention on SOC changes at a regional scale. Specifically, the improved approach utilizes regression kriging (RK) to estimate reference carbon stocks for the IPCC method. In our case study, we compared the results from the RI( method with a simple averaging (SA) method to derive the reference stocks as implemented in the Tier 2 IPCC approach, for a seven state area of the Midwestern United States. Using this improved method, we predict that eliminating tillage and retaining crop residues on all croplands of the study area would result in 11,735 Gg C yr(-1) sequestration for 20 years in the top 30 cm of the soil profile. Most cropland area would sequester 02-0.75 Mg C ha(-1) yr(-1). However, at a few places, the predicted rate of sequestration was more than 0.75 Mg C ha(-1) yr(-1), with an upper limit of 1.1 Mg C ha(-1) yr(-1). The highest rates of carbon accumulation were associated with favorable environmental conditions, such as lower slope positions and cold, temperate, moist climates. Validating predicted SOC change at 18 sites with varying soil types and environmental conditions showed that the RK approach to estimate reference carbon stocks decreased global prediction errors by 45% relative to the default reference values. The increase in prediction accuracy was due to using spatially varying SOC stocks rather than simple data averaging to derive reference SOC values. The uncertainty analysis demonstrated that there was more precision in the results from the RK approach in comparison to the results from the SA approach. These results suggest that improved geostatistical approach is a promising technique for improving soil carbon inventories that utilize the IPCC method, and will provide more precise results for informing public policy and management decisions while retaining ease of application.
