Developing historically-consistent and broadly-applicable monitoring, reporting, and verification system for quantifying forest change

Abstract

Given the increasing impacts of climate change and natural disturbances on forest ecosystems across the US, there is a need for monitoring systems that allow for accurate and rapid detection of historic and future changes in forest area and carbon stocks.  This collaborative project between UMN, USFS, and NASA is piloting a Monitoring, Reporting, and Verification (MRV) accounting system that could be used within the context of the National Greenhouse Gas Inventory baseline reporting to the UN Framework Convention on Climate Change. To accomplish this, baseline biomass density and historic data about forest change derived from Landsat and LIDAR information are being combined with USFS Forest Inventory and Analysis monitoring system to provide annual estimates of forest C stock and stock change from 1990 to present for several regions of the US. Products include maps and datalayers depicting forest change over broad geographic areas and a MRV accounting system for assesing forest area and carbon density.

Preliminary Results

Models of residence time (TRES) of downed woody debris (DWD) were developed to evaluate the impacts of future climate on this critical habitat and component of forest carbon stores. Results suggest that DWD in eastern US forests may decrease (i.e., more rapid decomposition) by as much as 13% over the next 200 years depending on various future climate change scenarios, altering C flux patterns in forest ecosystems.  Impacts of projected climate regimes (based on an ensemble of 17 CMIP5 models) was more severe for northern regions of the US, with decreases in residence time exceding 20 years for several northern tree species.  

Eastern US-wide assessments of patterns in plant functional traits and forest structure and biomass were conducted to evaluate the ability to predict broadscale relationships in carbon storage and mortality using trait data.  Patterns in the distribution of traits highlight higher levels of drought sensitivity for forests in upper New England and the Lake States and lower levels of biomass accretion in these regions relative to other portions of the eastern US.  As part of an effort to update estimates of carbon storage for US forests, models of belowground carbon pools that explicitly integrate temperature effects on belowground stocks were developed and indicate a decreasing level of belowground storage with projected increases in mean annual temperature by 2090. We also used Forest Inventory and Analysis data across the eastern US to examine forest C stocks and stock change by pool (e.g., aboveground live biomass) across general land-use categories (forest, agriculture, and developed).

Results indicate that the strongest sinks of forest carbon were in landscapes not completely dominated by forests, even when there was some loss of forests to agricultural/developed land uses likely reflecting the higher productivity rates of forests occurring on soils also suitable for agricultural production. Transfers of C from developed and agricultural land uses to forests from 2001-2012 represented a large net transfer of C (~40% of sink strength) that approached the same scale of net accumulation of C in forests remaining forests. Long-term land-use planning exercises and policy development should explicitly recognize the importance of minimizing forest to non-forest land use changes and the potential to expand forested habitats into areas significantly impacted by other land uses so as to maintain and enhance regional carbon sinks. In addition, the use of forest cover loss estimation based on remote sensing data may overestimate levels of forest carbon loss due to land-use conversion in areas where there is a high level of anthropogenic or natural disturbance to forest systems.