Forests are influenced by climate. They also influence climate because they remove carbon dioxide from the atmosphere as they grow. This project will better quantify both the influence of climate variation on tree growth and how much carbon is sequestered each year by forests by using two complementary data sets, the national forest inventory and tree-ring data. These data come from the U. S. Forest Service Forest Inventory and Analysis Program. This inventory program evaluates the current state and future trajectory of the nation's forests from measurements in a network of permanent sample plots. The plots are revisited and measurements are made once each decade in the western US. A spatially complete set of tree cores collected in the forest inventory plots contains information on the relationship between year-to-year variation in climate and tree growth, recorded in annual growth rings. The combined data will be used to study forest productivity across broad spatial scales and make predictions of ecosystem state, ecosystem services, and natural capital. This study will produce ecological forecasts relevant to society and informing strategies for climate adaptation. The project will also provide research experience and professional development for a post-doctoral fellow and undergraduates, open-source code for reproducible science, and outreach to schools and the public at the Laboratory of Tree-Ring Research at the University of Arizona. There is considerable uncertainty surrounding the future functioning of forest ecosystems, especially carbon sequestration, in the face of rising temperatures and evaporative demand. The focus of this project is the particularly vulnerable forest macrosystem of the interior western United States. Forest ecosystem functioning at this large scale will be analyzed by leveraging an existing, continental-scale ecological observatory network (the permanent sample plot network of the U. S. Forest Service's Forest Inventory and Analysis Program) and assimilating into it a new data stream with annual-resolution information on individual tree growth from increment cores collected in the same forest plot network. A hierarchical Bayesian multiple regression approach will be used to quantify the drivers of forest productivity across spatial scales and explain the effects of individual tree size, forest stand structure, geophysical conditions, disturbances, climate and their cross-scale interactions on tree-level and forest stand-level above-ground woody biomass production. Estimates of the effects of these drivers and their interactions will be used to make ecological forecasts of future tree- and stand-level productivity under various scenarios of stand density and stand size structure, including estimates of forecasting uncertainty. Data assimilation on this large scale will be supported by the cyberinfrastructure of NSF's CyVerse. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.