A spatial approach for multi-temporal estimation of forest growing stock volume and aboveground carbon pool. A case study in Tuscany (Italy)

Within the United Nations Framework Convention on Climate Change (UNFCCC) and the Paris Agreement’s Enhanced Transparency Framework, national greenhouse gas inventories are the key requirement to report GHG emissions by sources and removals by sinks, being the central element of transparency and understanding of the impact of climate mitigation. In this framework, quantitative information about forests plays a pivotal role in national and international monitoring programs and reporting activities. National forest inventories (NFI), complemented by wall-to-wall maps of forest variables, are usually the primary source of such information. NFIs are not designed as monitoring tools, since they are updated only every 5 or 10 years, but models can be used to produce annual carbon stock changes and fluxes. Following the approach of the FOR-EST model we present here a spatial approach to update growing stock volume (GSV) changes for years between national forest inventories, taking Tuscany (Italy) as a case study. The GSV update is mainly driven by the GSV current increment, predicted with forest types-specific growth models derived from yield tables. The spatial-explicit estimation of GSV is based on an initial GSV map, forest types, and forest disturbances maps. The spatial approach has provided comparable results with the original FOR-EST model, reaching a relative root mean square error (RMSE%) of 8.3% against the data reported by Italy under the UNFCCC. We also validated the results of our approach against an independent dataset of 342 circular plots distributed over the study area, measured between 2006 and 2019, reaching a mean RMSE% of 42 % and an R2 of 0.55 across the years.