Historical land-use change and ecosystem carbon balance in the continental United States

Release Date:

This article is part of the Spring 2019 issue of the Earth Science Matters Newsletter. 

Forests, grasslands, and shrublands absorb carbon dioxide from the atmosphere through photosynthesis by living plants and emit carbon dioxide back to the atmosphere through respiration by plants and soil decomposer organisms. These ecosystems act as net carbon sinks when they absorb more carbon dioxide in a year than they emit back to the atmosphere. This provides an important climate benefit by partially offsetting human emissions of greenhouse gases. Within the United States, terrestrial ecosystems offset approximately 12% of human carbon dioxide emissions per year. However, estimates of the size of the U.S. land carbon sink vary widely, in large part because of uncertainties surrounding the long-term impact of changes in land use and land cover.

This uncertainty exists because obtaining observational data of land use change over long time periods can be difficult, particularly across multiple ecosystem types over large geographic areas. Recently published research by USGS scientists used data derived from the Landsat satellite imagery archive to estimate the extent of land changes in the conterminous U.S. (i.e., excluding Alaska and Hawaii) and model the impact of these changes on terrestrial ecosystem carbon balance. Researchers used the Land Use and Carbon Scenario Simulator (LUCAS) modeling framework which includes a fully-coupled model of land-use and land-cover change and ecosystem carbon dynamics. The LUCAS model incorporates key uncertainties in land use change parameters and controls for the impact of climatic variability, which allowed researchers to isolate the effects of land change on ecosystem carbon dynamics.

The USGS researchers found that land use changes since the early 1970s resulted in large declines in agricultural and forested land area with a modest increase in grassland and shrubland land area in the conterminous U.S. The largest increase in any land cover class was driven by urbanization, with a gain of nearly 115,000 km2 of developed land between 1973 and 2010. The researchers used a carbon stock-flow model to estimate how the carbon stored in live biomass, dead wood, and soil changed over time and in response to land use change. Terrestrial ecosystems of the conterminous U.S. acted as a net carbon sink during most of the 37-year study period, sequestering an average of 254 Tg C yr-1 (Tg C yr-1 = one teragram of carbon per year = one million metric tons). There was no noticeable trend in the land carbon sink over the study period, which averaged 282 Tg C yr−1 in the 1970s, 224 Tg C yr−1 in the 1980s, 284 Tg C yr−1 in the 1990s, and 241 Tg C yr−1 in the 2000s. However, there was considerable inter-annual variability in the land carbon sink driven by short-term climatic variations, including at least one year (1988) where ecosystems were an overall net carbon source to the atmosphere (-152 Tg C yr-1) which was characterized by widespread drought and massive fires in the western U.S. Although forests were the largest contributor to the U.S. land carbon sink, forest carbon sequestration declined by 35% over the study period primarily because of ageing and a decline in forest area due to land use change.

graph of net biome productivity

In panel a, bars show the net biome productivity (NBP) for the conterminous United States by LULC class for the period 1973–2010. Points and error bars show the total NBP across the conterminous US with 95% Monte Carlo confidence intervals. The dotted line plots the mean annual NBP of 254 Tg C yr−1. The solid line plots the linear trend over the time series. In panel b, the solid line shows the annual net primary production (NPP) and the bars show the annual Palmer Drought Severity Index (PDSI). (Figure 5 in Sleeter et al., 2018)

(Credit: Figure 5 in Sleeter et al., 2018. Public domain.)

By incorporating map and satellite data accuracy assessments into their model, the USGS researchers determined that land use and land cover estimates prior to 1985 had roughly two times more uncertainty than estimates from more recent years. This higher level of uncertainty in land change data during the early years of the study contributed to a ∼16% margin of error in the annual carbon sink estimate prior to 1985. Uncertainty in NBP estimates due to LULC change was reduced by nearly 50% after 1985, largely due to increased spectral and spatial resolution, improved mapping capabilities, and increased data availability.

Information from this paper highlights that continued improvements in detection and attribution of the effects of land cover change on ecosystem carbon dynamics should further reduce uncertainties. This will result in improved science-based management recommendations to optimize the climate benefits provided by wildland ecosystems in the US.

The paper, “Effects of contemporary land-use and land-cover change on the carbon balance of terrestrial ecosystems in the United States”, was published in Environmental Research Letters. It is available at: http://iopscience.iop.org/article/10.1088/1748-9326/aab540/meta

<< Back to Spring 2019 Newsletter

Related Content

Filter Total Items: 1
Date published: November 30, 2018
Status: Active

The LUCAS Model

Our team is developing the Land Use and Carbon Scenario Simulator (LUCAS) model. LUCAS is a state-and-transition simulation model designed to track changes in land use, land cover, land management, and disturbance, and their impacts on ecosystem carbon storage and flux.