Climate mitigation and forest management require accurate information on carbon stocks, fluxes, and potential future sequestration potential. Previous large-scale estimates have substantial uncertainties arising from lack of data, heterogeneity of forest structure, and modeling limitations. However, recent local-to-regional studies suggest that combination of lidar-derived canopy height with an advanced 3-D ecosystem model that explicitly tracks vegetation height (i.e. Ecosystem Demography, ED) can reduce uncertainties and provide mapped estimates of these quantities at high-spatial resolution over policy relevant domains. Extending this approach to the global scale requires both a source of global lidar data height data and a global height structured ecosystem model. The NASA GEDI mission provides precise measurements of forest canopy height and vertical structure with great potential for global carbon cycle modelling. Here we present recent development and calibration of ED-global (v1.0) and its evaluation simulations against heterogeneous sources of satellite observations and field measurements. ED-global estimates of vegetation carbon stocks and fluxes, vegetation distribution and structure will be examined across various temporal and spatial scales from seasonal to inter-annual and also from grid cell to biome. The developed ED-global will serve as base model of NASA’s GEDI mission to answer the key science questions: What is the carbon balance of Earth’s forests? And how will the land surface mitigate atmospheric CO2 in the future?

Climate mitigation planning requires accurate information on forest carbon dynamics. Forest carbon monitoring and modeling systems need to step beyond the traditional Monitoring, Reporting, and Verification (MRV) framework of current forest cover and carbon stock. They should be able to project potential future carbon stocks with high accuracy and high spatial resolution over large policy-relevant spatial domains. Previous efforts have demonstrated the possibility and value of combining a process-based ecosystem model (Ecosystem Demography, ED), high-resolution (1-meter) lidar and NAIP data, field inventory data, and meteorology and soil properties in a prototype carbon monitoring and modeling system developed for the state of Maryland. Here we present recent work on expanding the Maryland prototype to a 10x larger domain, namely the Regional Greenhouse Gas Initiative (RGGI+) domain consisting of the states of Maryland, Delaware, Pennsylvania, New York, New Jersey, Rhode Island, Connecticut, Massachusetts, Vermont, New Hampshire, and Maine. The system expansion includes an updated version of the ED ecosystem model, improved initialization strategy, and expanded Cal/val approach. High-resolution wall-to-wall maps of current aboveground carbon, carbon sequestration potential, carbon sequestration potential gap, and time to reach sequestration potential are provided at 90m resolution across the RGGI+ domain. Total forest aboveground carbon sequestration potential gap is estimated to be over 2,300 Tg C for the RGGI+ region, about 1.5 times of contemporary aboveground carbon stock. States and counties exhibit variations in carbon sequestration potential gap, implying different policy planning for future afforestation/reforestation and forest conservation activities. Here we present the details of this new carbon monitoring and modeling system as well as regional results, including evaluations of our estimates against USFS Forest Inventory and Analysis (FIA) data, multiple wall-to-wall AGB maps, and state-wide and county-wide future carbon sequestration potential over time.