A METHODOLOGY FOR ASSESSING WATER TRANSIT TIME USING TEMPERATURE AS A NATURAL TRACER

This study presents a new methodology for assessing transit time of water based on a time variant cross-correlation analysis. The method is applied to two high-resolution temperature time series during a major flood event and its recession in the Santa-Fe River (North Central Florida) between the River Sink – where all the incoming surface water is drained by a sinkhole – and the River Rise – where the water resurges through a first magnitude karst spring. As a first step, the method requires selection of the relevant frequency domain for which temperature variations characterize the mass transfer between the two sites, which consist in application of a low pass filter with a cutoff frequency of 0.8 cycles per day to smooth the temperature dataset at the River Sink and the River Rise. Then, the two smoothed temperature time series are divided into sliding windows of a size such that the transit time is relatively stationary on the timescale of the window length. Time variant cross-correlation functions (ccf) are computed using the two smoothed temperature time series on each time window. The time lag for which each ccf is largest is used to assess the transit time of the water between the two monitored sites. Estimated residence time between the River Sink and the River Rise ranges from 8 hours during the flood to more than 4 days following the flood recession. This technique improves the ability to characterize the geometry of the main karst conduit, as well as the dynamics of recharge and transit through the aquifer, factors which are critical to the chemical and physical behavior of surface water-groundwater mixing in this system. This methodology has direct application to contaminant risk assessment, particularly with issues about the spread of contaminants between surface and groundwater bodies.