Assessing and Monitoring Spatial and Temporal Distributions of Ecosystem Carbon Storage and Changes in the United States

Land changes (land use and ecosystem disturbances) are the primary driver of stability and vulnerability of ecosystem carbon sequestration. Advances in remote sensing and modeling make it possible that carbon storage in relation to land changes can be assessed and monitored at the national and regional scales. Using remote sensing and modeling tools, the U.S. Geological Survey is conducting a national assessment to estimate spatial and temporal distributions of carbon storage in relation to land changes. The assessment covers all major ecosystems: forests, shrub and grasslands, croplands, wetlands, and aquatic systems. Recent land changes (baseline, 1992 to current) are mapped on an annual basis using Landsat imagery; future land changes (current to 2050) are modeled by incorporating IPCC socioeconomic storylines and climate change projections (three storylines and projections used: A1B, A2, and B1, each with multiple GCM runs). Carbon storage in, and transitions between, ecosystems are modeled and estimated annually using biogeochemical models, with the baseline and future potential land use changes and fire disturbances as the primary input. Effects of land changes and management activities are analyzed. A series of regional-scale maps and datasets are produced as deliverables of the assessment. The Great Plains region of the United States is the first region to complete for the assessment. The region encompasses 2.17 million square kilometers from eastern half of Montana south to Texas and east to Minnesota and Iowa. Changes in land use between 1992 and 2050 are pronounced for major ecosystems, including 7-16% gains in agriculture, 8-17% losses of grasslands and 18-19% losses of wetlands under A1B and A2 scenarios. More environmental oriented scenarios such as B1 will see gains in wetlands (15%) while holding areas of other ecosystems stable. For fire disturbances, number, size, and severity of large wildland fires in the region are highly variable, depending on IPCC scenarios and GCM runs, with variability of projected precipitation (and associated fuel moisture availability) particularly driving fire regime and fire emission distributions. Number of fires and areas burned may increase as much as 33% and 17% respectively compared to baseline conditions. Annual change of carbon stocks in terrestrial ecosystems depends strongly on existing soil organic carbon (SOC) level: soils with higher SOC levels tend to be C sources, and those with lower levels tend to be C sinks. Management practices (e.g., crop composition and rotation, drainage alternation, and conservation tillage) also play major roles driving the spatial and temporal patterns and changes of carbon fluxes in the region. And finally, lateral export of inorganic carbon and organic carbon via rivers and streams in the Great Plains watersheds is estimated at 1.3-1.5 and 0.25-0.55 g C m-2 yr-1 respectively. Putting these results together, we are able to present a complete carbon budget for all major ecosystems and controlling processes in the region. In addition, we will outline a carbon monitoring system centered on repeatable land change measurements using the methodology of the assessment.