Zampella, R. A., C. L. Dow, and J. F. Bunnell. 2001. Using reference sites and simple linear regression to estimate long-term water levels in Coastal Plain forests. Journal of the American Water Resources Association. 37:1189-1201. (Summary)
We used linear regression and long-term (10 years) stream-gaging and pitch pine lowland reference-site water-level data sets to estimate long-term hydroperiods at lowland test sites based on two short-term (two years) periods of record. Test sites had long-term records ranging from four to 10 years, which enabled us to validate our regression models. Although we found no overall significant difference between measured and predicted growing-season water levels, the analysis revealed a significant difference between months and test-sites. This difference increased through the growing season. Similar results were obtained using the stream-gaging site data, although the difference between measured and predicted growing-season water levels was greater with the stream-gaging data. Agreement between measured and predicted frequencies of near-surface saturation at test sites for each growing-season month was excellent.
