Surface Erosion Response to Climate Change in the Ouachita National Forest, USA

Sediment has been a major ongoing concern for the Ouachita National Forest in west-central Arkansas and eastern Oklahoma. Predicted climate change suggests that surface erosion risks may change in the future. To assess this risk for the Forest, we simulated how plausible changes in rainfall amount, rainfall intensity, and vegetation cover affect sediment risk under a future climate. Climate conditions were projected for the 2041-2060 period using an ensemble general circulation model and the A1B emissions storyline. The ensemble model averages the predictions of four general circulation models determined to be most appropriate for the U.S. Southeast. The A1B emissions storyline assumes an intermediate growth level in greenhouse gases and aerosols through the next century. Erosion risk was evaluated using a combination of a USLE-type model (for non-road areas) with the WEPP model (for unpaved roads). Annual and monthly precipitation, and mean monthly temperature were projected for the 2041-2060 period. The mean annual precipitation for the period was used to predict rainfall erosivity (the R-factor) for USLE modeling, while monthly precipitation and temperature data were used in WEPP modeling. The levels of three factors were varied to explore how sediment risks might respond to climate change: the R-factor, ground cover (USLE C-factor), and rainfall intensity (by varying the number of wet days per month in WEPP). The following factor levels were considered: ● R-factor: the predicted R-factors, and values 15% above and below the predicted values ● C-factor: the current C-factors and a 15% increase (i.e. decreased cover) ● Number of wet days: a 20% decrease from the current number and a 50% decrease Erosion risk was computed for all subwatersheds (typically 40-160 km2 in size) within the Forest (n = 190). The change in erosion risk was evaluated by comparing how the number of subwatersheds in the low, moderate, and high risk classes change relative to the current class number for each factor level combination. The predicted 2041-2060 climate shows an increase of 2.6 to 3.4 °C relative to the present. Mean annual precipitation decreases 0.9 to 7.0% on the western side of the Forest, but increases 0.4 to 2.6% on the eastern side. The predicted changes in R-factors (west: -1.8 to -13.7%; east: 0.6 to 5.3%) have little effect on the risk classifications unless a concomitant increase in C-factors occurs. With the C-factors unchanged, the future R-factors cause no change from the current numbers in each risk class. With 15% higher R-factors, 8% of the subwatersheds change from a lower to a higher risk class (worsen), while with 15% lower R-factors, 13% change from a higher to lower risk class (improve). However, if C-factors increase 15%, the number of the subwatersheds improving under the low-R level decreases 7%, while under the predicted and high-R levels, 7 and 14%, respectively, move into a higher risk class. Surprisingly, decreasing the number of wet days from 20 to 50% of the current number produces no appreciable difference in road erosion. These findings indicate that changes in vegetative cover may be just as important as rainfall changes in affecting future erosion risks on the Forest.