B23E-0251:
Importance of Past Human and Natural Disturbance in Present-Day Net Ecosystem Productivity

Tuesday, 16 December 2014
Benjamin S Felzer and Peter Phelps, Lehigh University, Bethlehem, PA, United States
Abstract:
Gridded datasets of Net Ecosystem Exchange derived from eddy covariance and remote sensing measurements provide a means of validating Net Ecosystem Productivity (NEP, opposite of NEE) from terrestrial ecosystem models. While most forested regions in the U.S. are observed to be moderate to strong carbon sinks, models not including human or natural disturbances will tend to be more carbon neutral, which is expected of mature ecosystems. We have developed the Terrestrial Ecosystems Model Hydro version (TEM-Hydro) to include both human and natural disturbances to compare against gridded NEP datasets. Human disturbances are based on the Hurtt et al. (2006) land use transition dataset and include transient agricultural (crops and pasture) conversion and abandonment and timber harvest. We include natural disturbances of storms and fires based on stochastic return intervals. Tropical storms and hurricane return intervals are based on Zheng et al. (2009) and occur only along the U.S. Atlantic and Gulf coasts. Fire return intervals are based on LANDFIRE Rapid Assessment Vegetation Models and vegetation types from the Hurtt dataset.

We are running three experiments with TEM-Hydro from 1700-2011 for the conterminous U.S.: potential vegetation (POT), human disturbance only (agriculture and timber harvest, LULC), and human plus natural disturbance (agriculture, timber harvest, storms, and fire, DISTURB). The goal is to compare our NEP values to those obtained by FLUXNET-MTE (Jung et al. 2009) from 1982-2008 and ECMOD (Xiao et al., 2008) from 2000-2006 for different plant functional types (PFTs) within the conterminous U.S. Preliminary results show that, for the entire U.S., potential vegetation yields an NEP of 10.8 gCm-2yr-1 vs 128.1 gCm-2yr-1 for LULC and 89.8 gCm-2yr-1 for DISTURB from 1982-2008. The effect of regrowth following agricultural and timber harvest disturbance therefore contributes substantially to the present-day carbon sink, while stochastic storms and fires have a negative effect on NEP. Even though the current NEP reflects the carbon uptake from regrowth, a full carbon accounting would also include the carbon released to the atmosphere during disturbance or carbon lost to decomposition of agricultural or timber products