Analyzing long-term hydrological impacts of forest disturbance and growth: a case study from Homochitto Watershed (MS)

The long term impacts of forest growth and disturbance on hydrological processes from a large watershed area (> several hundred km2) have been rarely studied due to data unavailability and complexity of methods. Combining time series of remotely sensing (1973-2006) and stream flow data (1938-2006), this study demonstrated efficient methods to quantify changes in forest landscape and to understand their long term impacts on hydrological processes. Time series data of precipitation and streamflows from two large sub-watershed areas (> 100 km2) were analyzed using spectral analysis and regression method. The spectral analysis method delineated the dominant signals from precipitation and streamflow, which corresponded to semi- and annual trends of the climatic system, and the tropical storm season observed in the southern states due to the El Nino Southern Oscillation (ENSO). Using this frequency content information, regression models were formulated and explained the overall trend and seasonal pattern of time series water yield data. Regression results showed the following: (1) approximately 30 % of total flow was due to baseflow contribution; (2) baseflow recharge increased with forest growth and maturity; (3) forest disturbance brought sudden increase in baseflow, but this change can be ameliorated with afforestration/refforestration efforts; (4) the increase in baseflow contribution was slightly greater at the downstream area, but the upstream area showed a slightly higher seasonal fluctuation. Overall hydrological responses examined by baseflow from two large areas with similar landscape characteristics were self-similar, regardless of differences in their geographical characteristics such as the drainage area size, and the degree of forest disturbance. The study was conducted at Homochitto Watershed in Missisippi, one of the most important timber producing regions of the United States