On The Effects Of Waves On The Wind Energy Input Into The Ekman Layer And The Ekman Transport

The dissertation is mainly involved by two parts. The first part puts insight intothe influences of ocean waves on the wind energy input into the Ekman layer, whilethe second part concentrates on the effects of wave-induced mass transport on theEkman transport.In the first part, based on the assumption of a depth-independent constant eddyviscosity coefficient, the expressions of the direct wind energy input and thewave-induced energy input to both steady state and non-steady state Ekman-Stokeslayer are deduced for the wave-modified energy balance equation of the Ekman layer,in which the effects of three wave-related terms (Stokes drift, wind input and wavedissipation) are all taken into consideration. It is for the first time that threewave-related terms’effects on the energetics of the Ekman-Stokes layer are given,which specifies the influence of the ocean waves in this respect. Meanwhile, for thefirst time the Stokes drift evaluated by the wave spectrum is employed to calculate theenergetics of the Ekman layer, thereby getting rid of the limitations of themonochromatic wavefield assumption. Besides, the expressions of wind energy inputare also presented for the condition without wave dissipation and that with Stokesdrift only.Based on the wave-number spectrum of Donelan and Pierson (1987) for a fully developed wind-generated sea and the spectrum data from WAVEWATCH III ideal tests, which are driven by time-independent wind speeds, the contributions of the threewave-related terms to the wind energy input into the Ekman layer are discussed andproved to be of significant influences respectively. Stokes drift is introduced into theenergy balance equation of classical Ekman layer, which causes the reduction in thedirect wind energy input and provokes a new energy term, that is, the wave-inducedenergy input; then the reduction of wave-induced wind stress term in the energybalance equation conduces to the decrease in both the direct wind and wave-inducedenergy input; furthermore the introduction of wave-induced momentum transfer fromwave to the mean flow due to dissipation of wave energy leads to the increase in allenergy input terms. The results also show that the ratio of the wave-induced to thedirect energy input into the Ekman layer goes up with the ascension of the wind speed and the latitude, and two energy input terms are of the same order of magnitude in thehigh latitudes with moderate wind speed.The global ocean wave hindcast model is established based on WAVEWATCHIII with CCMP wind data providing estimation of wave parameters and wavespectrum data in area to the south of30°S latitude and the whole Pacific ocean withgrid resolution of2degrees. The wave spectrum data can be used to calculate Stokesdrift and some other wave-related terms, which are essential variables for study on theenergy input to the non-steady state Ekman layer and the wave-induced masstransport. Comparison with in-situ buoy data from National Oceanographic DataCenter validates that the model performs well in representing the Stokes drift field.Using global ocean wave model and the CCMP wind data,23-year andseasonally averaged wind stress, Ekman depth, Stokes drift, wave-induced momentumtransfer and wave-induced wind stress reduction in the South Ocean are presented.Besides, the23-year averaged direct wind, wave-induced and total energy input to theEkman layer in the Southern Ocean are estimated and the results show that the energyinputs are the strongest in the Antarctic Circumpolar Current (ACC) area. If threewave-related terms are included, the ratio of integral wave-induced to total energyinput in the Ekman layer is15%. Total energy input is about403GW, while the directwind and wave-induced energy inputs are343GW and60GW, respectively.Meanwhile, the long-term trends of the energy inputs into Ekman layer in the ACCarea are discussed, and the rising trend of the wind stress in this area during the past23years may contribute to the ascendant of the energy inputs.In an attempt to compare with the results of recent studies, this dissertation alsoadopts the same data and methods as in previous research and re-estimates energyinput into Ekman layer. It turns out that there is underestimation of direct wind energyinput and overestimation of wave-induced energy input for former research resultsthat employ the ECMWF wind and wave data and calculate Stokes drift with bulkwave parameters.In the second part, wave-induced mass transport in the Pacific and SouthernOcean are re-estimated on the basis of Stokes drift field calculated by wave spectrumdata output from the WAVEWATCH wave model. The wave-induced wind stressreduction term is introduced into the expression of classical Ekman transport, which isdefined as the wave-effected Ekman transport. In addition, presented are the classicaland wave-effected Ekman transports in the Pacific and Southern Ocean. The comparison with works conducted by McWillianms&Restrepo (1999) and Polton etal.(2005) shows that their estimation about the Stokes drift calculated by PMspectrum leads to overestimation of the ratio of wave-induced transport to Ekmantransport, on the contrast, the ratio in the westerlies is merely0.2~0.3on condition ofStokes drift calculated by model spectrum. If the wave-induced wind stress reductionterm is included, the ratio in the westerlies is as high as0.4~0.5, which means that thewave-induced mass transport significantly impacts the total mass transport in theupper ocean.