Reconstructing Multi-Century Streamflow Records in the Mobile-Tensaw Delta
Across the Southeastern United States (SEUS) growing populations are increasing demand on water resources and infrastructure. Understanding the long-term natural flow regime of rivers is critical to developing accurate models of water level variability needed for appropriate water resource management. Insufficient hydroclimate records fail to accurately capture the frequency of severe droughts or to document long-term monotonic changes in climate, like increased aridity, humidity, or changes in consumption (Crockett et al., 2010). We used new and existing tree-ring chronologies to reconstruct May-August discharge for the Alabama River during the period 900-2011 CE in order to place the period of instrumental flows (since 1931 CE) into historical context.A nested principal components regression model was used to reconstruct streamflow, maximizing the use of chronologies with varying time coverage in the network. The regression model applied utilized the mean index chronology as the predictor for the climate-variable that most influences tree growth at our site. The modeled streamflow estimates indicate that streamflow conditions of the instrumental period do not sufficiently represent the full range of Alabama River flow variability beyond the observational period. Although extreme hydroclimate variability is present in the gage record, the tree-ring record suggests that the intensity and duration of flood and drought events that occurred during the 1500s and 1700s was far more severe. These findings imply that basing future water policy on water availability witnessed during the instrumental period could result in devastating water shortages if droughts as intense as those in the 16th and 18th century were to occur in modern times.