论文部分内容阅读
To experimentally estimate heat fluxes,the eddy covariance method is applied using time series of wind speed components u’,v’,and w’ and virtual temperature fluctuations Tv’ by sonic anemometers.A recognized issue with determining these fluxes is a systematic overestimation of the covariance during high wind conditions.In the last few years,many papers have addressed the limitation of the Monin-Obukhov theory in the surface layer,especially in marine environment.Factors taken into considerations are non-local effects i.e.entrainment,advection,waves,or the presence of sink and sources of heat within the surface layer,i.e.sea spray; the last two factors are more important during high wind speed.The issue of the role of sea spray in the transfer of latent and sensible heat was reconsidered by re-analysing data from the HEXOS experiment at a research platform in the North Sea.Sea spray was found to contribute up to 40% of the total latent flux for wind speed between 15ms-1and 18ms-1.Here,we present results from measurements performed on an Antarctica ice sheet,at a coastal site,and offshore.We believe that high heat fluxes are caused by a combination of instrumental errors and physical effects and we discuss the different cases where the increase of heat fluxes is more or less evident and whether and how it would be possible to quantify the overestimation by sonic anemometers.Estimates of heat fluxes,in Antarctica,during high-speed katabatic regime,show negative values up to - 300Wm-2.Negative fluxes of such a magnitude immediately raise doubts regarding the accuracy of measurements,however,they are supported by observed temperature gradients,although fluxes estimated using flux-profile relationship based method are not larger than - 100Wm-2.During high wind,the applicability of flux-profile relationship might be limited by the presence of blowing snow,which introduces sink and source of heat in the surface layer likewise sea spray.Furthermore,mixing of warmer air flowing above the ice during katabatic events,and the increase of velocity fluctuations with increasing wind speed,also support the reliability of large negative heat fluxes.We addressed the effect of sea spray considering the basic issue of the correlation coefficient between humidity q’ and temperature T’ fluctuations defined as RqT=/σqσT,where is the q,T mean covariance,and σq and σT are their standard deviations.Since q and T are scalars,they are transported by turbulence,therefore their correlation should be RqT=1.However,RqT seldom reaches such a value in the surface layer due likely to non-local effets; Measurements of concentration of marine aerosols were collected at a coastal site during high wind conditions,and we found a high concentration of sea spray when 0.4< RqT<0.8,which could explain the low correlation between q’and T’ and therefore the non applicability of the flux-profile relationship.