ESTIMATING STEADY AND UNSTEADY TRANSPORT IN THE KARST SYSTEMS USING QUANTITATIVE TRACER TESTS AND CORRELATION ANALYSIS

The quantitative interpretation of tracer tests provides an estimate of the transport parameters in the karst systems, which are valid for specific flow conditions. The repeated tracer experiments under steady and unsteady flow conditions furnish a more complete understanding of the transport dynamics, but in remote regions, that solution would lead to very expensive and difficult logistic. This study correlated the parameters obtained by quantitative tracer tests (under quasi-steady conditions) with delays measured by cross-correlation analysis of precipitation against flow rates in springs. The objective was to evaluate an alternative method to characterize the transport during unsteady flow.

The data were obtained in four karst systems located in the southern mountains of São Paulo State (Southeastern Brazil). The distances between the injection points in the sinkholes and the springs range from 2.85 to 6.8 km; the time for the peak tracer concentration (TP) range from 16.2 to 68.8 h; the time for the first arrival of the tracer (FA) vary from 12.7 to 49.3 h and the delays vary from 4 to 18 h (measured during one hydrologic year 2009-2010).

A simple linear regression of the distance against delay time (n=6) shows a coefficient (R²) equal to 0.51, shifting to 0.67 if only the main and more distant injection points are considered (n=4). The linear dependence between distance and response time indicates that the systems are not dominated by phreatic flow, and the increase of the R² shows that delays are more correlated with the pulses injected in the more distant and efficient sinkholes. The correlations between TP and distance show R² = 0.83 (n=6) and R²= 0.75 (n=4) and between FA and the distance show R² = 0.92 (n=6) and R² = 0.90 (n=4). The correlations of the delays against TP and FA were R² = 0.44 (n=6), R² = 0.71 (n=4) and R² = 0.41 (n=6), R² = 0.71 (n=4), respectively.

Despite the small number of samples, the coherence of these results indicates that the delay between rainfall and discharge is a reasonable estimator of particle travel times during unsteady flow in karst systems not dominated by phreatic conditions. This procedure can be used as an auxiliary technique to evaluate the dynamics of contaminant transport in environmental studies.

Research sponsored by FAPESP (Proc. 2009/05115-5).