REVISITING SPRING CLASSIFICATION USING CONTINUOUS DATA LOGGERS (Invited Presentation)

A monitoring study of multiple karst springs in central Pennsylvania in 1968 yielded insight into relative contributions of matrix and conduit flow as indicated by seasonal variability in spring chemistry and temperature (Shuster and White, 1971). Some springs showed little temperature variation and higher dissolved ions (indicative of matrix flow contribution) while others showed seasonal temperature variations and more variable dissolved ions (indicative of more contribution from conduits). Newer studies have begun to incorporate continuous high temporal resolution monitoring of temperature, pH, conductivity and storm water samples. We returned to seven springs from the original study area in central PA and applied high resolution monitoring. In addition, four urban karst springs close to Philadelphia were examined. The continuous temperature loggers more readily identified seasonal and storm-scale variations in springs. Some springs previously identified as having little seasonal variation showed a small increase, but delayed from the peak summer temperatures. This delay in response indicates exchange of heat with the surrounding rock along the travel path. Comparing temperature profiles from 1968 to the present, most springs showed an elevation in summer temperatures of 1-2 degrees C but not in fall and winter. In addition, most of the springs also showed storm response in temperature, although the degree of flashiness varied even with adjacent springs and across seasons. Fewer storms were observed in the urban karst springs, possibly due to capture of water by storm sewers. Because temperature responses depend on conduit network geometry and ability to exchange heat into the surrounding rock, these temperature patterns indicate that flow path length and recharge pattern varies both spatially and temporally at a single location. These data provide further evidence that storage and recharge vary under different storm events. Because source area protection depends on recharge areas and travel times, continuous monitoring is needed to account for such variation.