Comparing observation-based estimates of active-layer thickness with Earth system model predictions
The depth of seasonally thawed soils (active layer) in the arctic environment determines the quantity of soil carbon emissions that may affect the global climate. Emissions increase as permafrost beneath the active layer melts and becomes part of the active layer. The accuracy of spatial representations of the active-layer thickness in Earth system models will limit the validity of predictions of future warming due to these emissions.
EVS assistant geospatial modeler Umakant Mishra and his colleague Bill Riley from Lawrence Berkeley National Laboratory coupled geospatially referenced soil profile observations with environmental variables such as topographic attributes, climate, and land cover types to estimate the variability of active-layer thickness across Alaska at fine spatial resolution. The results ranged from 0.14 to 0.93 m, with a spatial average of 0.46 m and a 30% coefficient of variability. Surface air temperature, land cover type, and slope angle were the primary environmental controllers of the spatial variability.
Comparison of the estimates to the Coupled Model Intercomparison Project Phase 5 (CMIP5) Earth system model predictions used in the current IPCC fifth assessment showed that model projections of active-layer thickness for Alaska had lower spatial variability but substantially larger prediction errors than the observation-based estimates. In addition, large inter-quartile ranges in predicted active-layer thickness (0.35-4.4 m) were found in CMIP5 projections, with some model predictions substantially greater than the geospatial estimates.
This comparison calls attention to the need for better process and spatial heterogeneity representations in Earth system models to enable more realistic estimates of regional-scale active-layer thickness. Such an improvement could decrease uncertainty in predicting the permafrost carbon feedback to climate under future warming scenarios.
This study has been accepted for publication in the Soil Science Society of America Journal – “Active-layer thicknesses across Alaska: Comparing observation-based estimates with CMIP5 Earth System Model predictions,” by U. Mishra and W.J. Riley.