Measurement uncertainty in gas exchange coefficients

A technique to measure the gas exchange coefficient (transfer velocity) from a three-dimensional unsteady tracer cloud on large water bodies is described. Arguments based on first principles are presented in support of the technique. For this reason, the measurement uncertainty is apparent and can be quantified analytically. The dual gas exchange measurement technique was tested and results are discussed through computational experiments that simulate a tracer cloud and incorporate random sampling. Measurements of gas exchange coefficient are undertaken in Chequamegon Bay of Lake Superior. The intensive sampling regression technique developed was used in these measurements, and a minimum percent uncertainty was used to determine an optimum analysis period. Results were determined with a 4 to 9 hour sampling period, with no preliminary mixing period.