Impact of non-idealities in gas-tracer tests on the estimation of reaeration, respiration, and photosynthesis rates in streams.

Estimating respiration and photosynthesis rates in streams usually requires good knowledge of reaeration at the given locations. For this purpose, gas-tracer tests can be conducted, and reaeration rate coefficients are determined from the decrease in gas concentration along the river stretch. The typical procedure for analysis of such tests is based on simplifying assumptions, as it neglects dispersion altogether and does not consider possible fluctuations and trends in the input signal. We mathematically derive the influence of these non-idealities on estimated reaeration rates and how they are propagated onto the evaluation of aerobic respiration and photosynthesis rates from oxygen monitoring. We apply the approach to field data obtained from a gas-tracer test using propane in a second-order stream in Southwest Germany. We calculate the reaeration rate coefficients accounting for dispersion as well as trends and uncertainty in the input signals and compare them to the standard approach. We show that neglecting dispersion significantly underestimates reaeration, and results between sections cannot be compared if trends in the input signal of the gas tracer are disregarded. Using time series of dissolved oxygen and the various estimates of reaeration, we infer respiration and photosynthesis rates for the same stream section, demonstrating that the bias and uncertainty of reaeration using the different approaches significantly affects the calculation of metabolic rates.