A Study On The Mesoscale Eddies In The Northwestern Pacific Ocean

The northwestern Pacific Ocean is one of the regions with most complexcirculation system in the world, and the mesoscale eddies are also very active in thisregion. As an important physical process in the northwestern Pacific Ocean, themesoscale eddies have become the focus of research in recent years. The mesoscaleeddies play very important roles in ocean dynamics and their transports of heat, salt,and energy, as well as other biological and chemical processes, which further impacton the circulation in this region, and the vertical and horizontal distributions oftemperature, salt, chlorophyll and so on. Therefore, research on the mesoscale eddiesin the northwestern Pacific Ocean has important scientific significance and practicalvalue for both ocean dynamics theory and military defense.First, using over15years’ satellite altimetric data, we adopt three distinctautomated methods for eddy identification （SSHA, Okubo-Weiss; OW, andWinding-Angle; WA） to analyze the statistical characteristics of mesoscale eddies’properties in the northwestern Pacific Ocean, and the comparison among the resultsare demonstrated. Then, combining Argo data with altimeter data, thethree-dimensional structures of eddies are reconstructed. Finally, the seasonal andinterannual variability of the thermohaline structures of eddy are further discussed.The structure and main results of this dissertation are shown as follows:The SSHA-based method is first used to identify mesoscale eddies in thenorthwestern Pacific Ocean （115°-140°E,5°-30°N）. The properties of the identifiededdies show that most of the eddies are located north of16°N where the CE and AEsdistribute evenly. The majority of eddies have mean lifespan less than12weeks withradius larger than100km. There are about51eddies generated on average per year.And the number of AEs is a bit more than that of CE in most years.Furthermore, the OW method is applied to detect eddies in the eddy-rich region（122°-148°E,12°-28°N）. The results show that the eddy occurrence frequency is largebetween17oN and24oN. Most of eddies have radii between40km and70km, and the mean radius and lifespan of the CE （AEs） are59.4km and8.5weeks （57.7km and8.9weeks）, respectively. Overall, an eddy with larger radius will have larger energybut smaller intensity.Mesoscale eddy properties in the northwestern subtropical Pacific Ocean（122°-170°E,12°-28°N） are investigated by using the third method WA. Eddyoccurrence frequency and kinetic energy are prevailingly high in the SubtropicalCountercurrent （STCC） zonal band between19oN and26oN, which is further elevatednear the Luzon-Taiwan coast. A general superiority of AE is observed at mostlatitudes except between19°N and22°N where CE number is larger. The majority ofeddies have mean lifespan less than10weeks with radius80-180km. Aftergeneration, most eddies propagate westward with a mean speed of7.2cm s^-1 and thendeflect northward following the Kuroshio along the Luzon-Taiwan coast.Through comparison of above three distinct eddy detection methods, we can seethat there are similarities in some aspects. For example, most of the eddies havelifespan less than10weeks, and all the methods successfully identify the eddies withlifespan longer than24weeks. For the eddies with different lifespans, AEs are a bitmore than CE. The eddies move westward after generation and propagate northwardfollowing the Kuroshio when approaching the west boundary. Although all three eddydetection methods are on the basis of physical or geometric criteria of the SSHanomaly fields with automated algorithms, there are some differences incharacteristics of the detected eddy. The mean radius of eddies detected by the OWmethod is less than60km and shows great meridional difference, which detected bythe SSHA and WA methods is larger than100km, and the zonal mean radius showslittle meridional variation for the WA method. For the long-lived eddies, theevolutions of the eddy properties detected by the OW method are more irregularbecause of the higher excess of detection rate, whereas the results from the WAmethod are more regular. Moreover, the eddy mean propagation speed derived fromthe OW method is only half of that from the WA method which is much closer to theresults of the previous studies. Thus, we consider that the WA method is more accuracy and the associated results are more reliable.For tracking and comparing the difference of the three-dimensional eddystructures in the westeard propagation, the study region （122°-170°E,18°-26°N） isdivided into five continuous subregions, and the three-dimensional eddy structures areexplored with composite eddy images in each subregion constructed by surfacingArgo temperature/salinity data into altimeter-detected eddy areas. The compositededdies could induce great thermocline vertical movement. Due to the existence of theNorth Pacific Subtropical Mode Waters （STMW） in the main thermocline, theupwelling （downwelling） of water mass induced by CEs （AEs） strain （compress） thethick STMW and make the double-core temperature anomaly structures more （less）discernable. Because the vertical distribution of the North Pacific Tropical Water（NPTW） and the North Pacific Intermediate Water （NPIW）, salinity anomaly featuresa sandwich-like pattern which is more evident in AE images. Also revealed is thesignificant structure difference in these five subregions. Eddies are greatly intensifiedas approaching the western boundary, inducing larger temperature, salinity, andgeostrophic current velocity anomalies and influencing deeper ocean. Moreover, ouranalysis indicates that the subtropical front in the study region could affect thecomposite temperature/salinity anomaly structures. However, comparing with ouraddressed large zonal variations, the meridional difference is secondary.Furthermore, we define the strong/weak seasons （years） of eddy activity basedon seasonal （interannual） variability of the eddy EKE, and analyze and compare theseasonal and interannual variability of the thermohaline structures of eddies bycomposing CEs and AEs, respectively. The seasonal variabilities of the EKE of CEsand AEs are synchronous: the eddy EKE is strong in spring-summer （April-June） andweak in fall-winter （November-January）. The temperature and salinity anomaliesinduced by CEs and AEs are mainly located in the upper100dbar. The interannualvariability of CEs and AEs are out of sync: the EKE of CEs is strong between January2003and September2006but weak between September2006and December2011;whereas the EKE of AEs is strong in2003-2007and weak in2009-2011. The temperature and geostrophic current anomalies induced by the eddies and theirinfluencing depth in the strong years are all greater than those in the weak years. Thedifferences of salinity anomalies induced by CEs and AEs between the strong yearsand weak years are mainly reflected in the upper200dbar due to the significant zonaldifferences of the sea surface salinity.