Drag and Bulk Transfer Coefficients Over Water Surfaces in Light Winds

Abstract

The drag coefficient (C D), experimentally determined from observed wind speed and surface stress, has been reported to increase in the low wind-speed range (< 3 m s −1) as wind speed becomes smaller. However, until now, the exact causes for its occurrence have not been determined. Here, possible causes for increased C D values in near-calm conditions are examined using high quality datasets selected from three-year continuous measurements obtained from the centre of Lake Kasumigaura, the second largest lake in Japan. Based on our analysis, suggested causes including (i) measurement errors, (ii) lake currents, (iii) capillary waves, (iv) the possibility of a measurement height within the interfacial/transition sublayer, and (v) a possible mismatch in the representative time scale used for mean and covariance averaging, are not considered major factors. The use of vector-averaged, instead of scalar-averaged, wind speeds and the presence of waves only partially explain the increase in C D under light winds. A small increase in turbulence kinetic energy due to buoyant production at low wind speeds is identified as the likely major cause for this increase in C D in the unstable atmosphere dominant over inland water surfaces.

Improving surface heat flux estimation for a large lake through model optimization and two‐point calibration: The case of Lake Geneva

“…a). Under high wind regimes, the calibrated value (~2 × 10 −3 ) and the almost linear variation of the transfer coefficients are in agreement with reported measurements (Graf and Prost ; Merzi and Graf ; Mahrt et al ; Babanin and Makin ; Toffoli et al ; Wei et al ). Although the transfer coefficient values are consistent with other studies, there is a shift in the wind speed under which these minima are observed.…”

“…To verify the calibration factors, we compare the shape and range of these curves with some measurements taken over water. The general form of theses curves is similar to the measured values over inland and open waters (e.g., Wüest and Lorke ; Wei et al ). As we obtained lower values for both C m 2 and C q 1 than Zeng et al (), the calibrated transport coefficients are smaller (higher) at low (high) winds than the uncalibrated coefficients in Fig.…”

“…The minimum values of the transfer coefficients (~10 −3 ) were obtained at a low wind speed of ~2 ms −1 (Fig. ), whereas the minima reported in the literature are, for example, at ~4 ms −1 (Mahrt et al ; Lin et al ) or ~5 ms −1 (Wüest and Lorke ; Wei et al ).…”

“…Higher coefficient values (~2 × 10 −2 ) are reported mainly for ocean experiments (Geernaert et al ; Bradley et al ) while lower values (~2 × 10 −3 ) were observed over shallow coastal waters (Mahrt et al ) and estuaries (Lin et al ). Wei et al () found a high drag coefficient value of ~10 −2 at a wind speed of ~1 ms −1 for Lake Kasumigaura (Japan), while a value of ~2.5 × 10 −3 was measured for Lake Neuchatel (Switzerland) for the same wind speed (Simon ). For Lake Kasumigaura, however, the humidity transfer coefficient was ~3 × 10 −3 at 1 ms −1 wind speed, which is closer to our estimated values (Fig.…”

An Experimental Investigation of Kinetic Fractionation of Open‐Water Evaporation Over a Large Lake

“…However, the kinetic factor is not very sensitive to wind speed, varying only slightly, from 6.9‰ at 10 m s −1 to 6.0‰ at 3 m s −1 , according to the theory of Merlivat and Jouzel (). The Dalton number of evaporation, a measure of the intensity of turbulent exchange of water vapor between the water surface and the overlaying air layer, does not seem to depend on lake size either (Wei et al, ; Xiao et al, ). Therefore, factors other than levels of turbulence, such as the compensating effect noted below, may be a more plausible explanation for why the LK ε k works in some situations but not in others.…”