Gas transfer velocities measured at low wind speed over a lake

The relationship between gas transfer velocity and wind speed was evaluated at low wind speeds by quantifying the rate of evasion of the deliberate tracer, SF6, from a small oligotrophic lake. Several possible relationships between gas transfer velocity and low wind speed were evaluated by using 1-min-averaged wind speeds as a measure of the instantaneous wind speed values. Gas transfer velocities in this data set can be estimated virtually equally well by assuming any of three widely used relationships between k600 and winds referenced to 10-m height, U10: (1) a bilinear dependence with a break in the slope at ;3.7 m s21, which resulted in the best fit; (2) a power dependence; and (3) a constant transfer velocity for U10 , ;3.7 m s21, with a linear dependence on wind speed at higher wind speeds. The lack of a unique relationship between transfer velocity and wind speed at low wind speeds suggests that other processes, such as convective cooling, contribute significantly to gas exchange when the wind speeds are low. All three proposed relationships clearly show a strong dependence on wind for winds .3.7 m s21 which, coupled with the typical variability in instantaneous wind speeds observed in the field, leads to average transfer velocity estimates that are higher than those predicted for steady wind trends. The transfer velocities predicted by the bilinear steady wind relationship for U10 , ;3.7 m s21 are virtually identical to the theoretical predictions for transfer across a smooth surface. The gas exchange of slightly soluble gases between surface waters and the atmosphere is driven by near–surface water turbulence (Donelan 1990). Wind is responsible for a significant portion of the turbulence in most surface waters and is quite easy to measure. As a consequence, gas exchange is often parameterized as a function of wind speed. When wind speeds are low, the relative contribution of wind to the overall turbulence in surface waters diminishes, and other factors, such as convective cooling (Crill et al. 1988) and chemical enhancement (for reactive gases) (Wanninkhof and Knox 1996), are thought to become increasingly important. As a consequence, the primary controls on gas transfer velocity are not well understood when wind speeds are low. This problem is exacerbated by the limited number of gas transfer velocity estimates that have been made in the field during periods of low wind. Among the only measurements are those of Emerson et al. (1973), Wanninkhof et al. (1985), Clark et al. (1995), and Cole and Cariaco (1998). The quantification of gas transfer at low wind speeds is of importance for biogeochemical mass balance studies that involve gaseous species and for estimating volatile pollutant transfer in lakes and inland waters, where low wind speeds are common. In the open ocean, gas transfer at low wind 1 Present address: Woods Hole Field Center, U.S. Geological Survey, 384 Woods Hole Road, Woods Hole, Massachusetts 02543 ( [email protected]).