PERFORMING A SENSITIVITY ANALYSIS ON VERTICAL FLUX ESTIMATES BASED ON SIMULATED STREAMBED TEMPERATURE TIME SERIES DATA

Using the temperature time series method to measure vertical flux of the hyporheic exchange in streams has gained extensive use in recent years. This is due to advancements in field instrumentation and computational tools used to process these data. In this study, we analyze the uncertainty associated in these methods by performing a sensitivity analysis to, (1) determine how acute changes in the temperature gradient may affect vertical flux estimates produced using the temperature time series method, and (2) analyze the change between measured and observed vertical flux estimates with respect to temperature sensor pair spacing and their center of depth value. To perform this analysis we used numerical models created with COMSOL Multiphysics and compared the output with the vertical flux estimates they generated using an analytical solution. To investigate aim one, we held the vertical flux constant and introduced various temperature fluctuations in the temperature time series data to determine at what time interval accurate vertical flux estimates are captured. This will enable users to determine the length of time sensor deployment should be maintained, in the occurrence of rapid atmospheric temperature shifts, to calculate accurate vertical flux values. For aim two we created a contour plot, using a root-mean-square error calculation, to present the amount of error associated with a sensor pair’s spacing verses its depth. From this plot, users can better determine at what subsurface depth and sensor pairing length consistent and accurate results are likely to be produced. Based on preliminary results, researchers may be able to use these findings to improve sensor probe deployment out in the field in order to reduce error in determining vertical flux in the hyporheic zone of a stream.