Investigating Consequences of Climate Change - Temporal and Spatial Changes of Precipitation, Runoff and Erosion and the Impact on Land Cover (Walnut Gulch Watershed, Arizona)

The spatio-temporal changes of precipitation as consequences of climate change may lead to the changes of soil erosion, runoff, and land cover in semi-arid rangelands. Walnut Gulch Experimental Watershed in Arizona, with its long-term observations and densely distributed rain gages and flumes, provides opportunity to quantify the spatial and temporal dynamics of precipitation events and the responses of soil erosion and runoff. The different vegetation covers between subwatersheds within Walnut Gulch Watershed makes it possible to access the effects of climate change and management on vegetation cover. Hydrologic modeling techniques provide insights for the evaluation of climate and management consequences on a natural system; however, the capability of the model to represent climate change and simulate hydrologic changes under climate change needs to be tested.;First the temporal variability of precipitation event was analyzed with a long-term moving climate period. The procedure utilizes the AR (1) time series autocorrelation model to adjust the time series dependence. The analog-to-digital transition effects were analyzed and adjusted with a one year overlap of digital and analog climate. The yearly time series analysis was conducted in parallel to compare with the long-term moving climate period analysis. The significant decreasing trends found in event amount, peak 30 minutes rainfall intensity (I30), event duration and rainfall energy suggested the climate is becoming dryer. The spatial variations of rainfall events of different scale within Walnut Gulch Watershed are quantified. Uniform distributions of long-term temporal precipitation between Lucky Hill, Kendall, and Tombstone sites suggests that the likelihood of climate change to explain the vegetation cover differences may be small. In addition, the variation of starting time of rainfall events at this scale confirms the highly heterogeneous nature of precipitation events. Within the small watershed scale, apart from spatial variation of rainfall, the variation of precipitation events also reflects the measurement error that might have been underestimated in the previous researches.;The rainfall-runoff relationships are evaluated within Lucky Hill watershed, both through assessing the aggregated annual amount and through the starting time of events. The rainfall events with I30>=22 mm/h are most likely to cause runoff, and the power law equation may better represent rainfall-runoff amount relationships within watershed. The experiment of event analysis indicated the errors in time measurement and missed recording of precipitation events of both flume and rain gages, and the possibility of spatial variation of precipitation events within small watershed.;WEPP simulations under measured climate were calibrated and validated with measured runoff and sediment data, and the results confirms the capability of using WEPP for long-term soil erosion and runoff simulation. The comparisons were made among WEPP simulations of sediment yield, runoff and biomass with long-term observed climate and CLIGEN generated climate. The return periods of sediment yield, runoff, peak rate, precipitation amount simulated from observed climate and CLIGEN generated climate were analyzed. The results supported the capability of using CLIGEN and WEPP to represent precipitation change and the corresponding changes of peak rate and sediment yield when combined with measured climate. In WEPP, the vegetation cover is more sensitive to rangeland management than climate change. This may imply the different vegetation cover between Kendall and Lucky Hill is more likely to be explained by different management instead of climate change.