透過您的圖書館登入
IP:18.188.108.54
  • 期刊
  • OpenAccess

Relationship Between the Dispersion of Thermal Effluent and the Tidal Current in the Waters Near the Outlet of the Third Nuclear Power Plant in Southern Taiwan

臺灣南部核三廠排放口附近海域溫排水擴散與潮流之關系研究

若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


利用超音波直接評估魚羣豐度時,當魚羣受內外因素影響而使成羣結構產生變化,會導致評估誤差,因此,為提高評估魚羣豐度之準確性,直接由聲探囘訊特性來區別不同單羣結構魚羣的反射信號實有必要。因此本實驗由44羣日本鮆(Engraulis joponica)魚羣之成對立體照片組,分析魚羣內個體間之相對位置後,利用Tylar(1961)及Clark and Evans(1954)法逐一判別其單羣結構是屬規律性結構或塊狀結構,然後根據每羣魚羣的連續超音波反射訊號,利用信號處理法分別篩選出具有單羣結構意義的7項反射性變數値,再以統計之逐步判別分析法及分類鑑別分析法篩選出區分不同結構魚羣的最佳獨立變數項組合,並建立一直接判別單羣結構的模式,探討直接由日本鮆魚羣的超音波囘訊中監測其成羣結構變化以提高評估準確度之可行性。結果摘要如下:(1)7項具有單羣結構意義的反射特性變數以逐步判別分析法篩選結果,發現由連續脈波的囘訊準位變化、覆蓋脈波數、脈波幅寬變化等3項獨立特性變數之組合來判釋魚羣的成羣結構屬性,具有最佳之區分效果。(2)規律性結構魚羣的連續超音波反射具有波峰數較少、囘訊準位及相位變化較穩定的特性;而塊狀結構魚羣則呈現多峰型且幅寬、囘訊準位及相位變化較大的反射特性。(3)由上述三項區分效果最佳的反射特性變數項,利用分類鑑別分析法依照不同成羣結構魚羣的超音波囘訊差異所建立的鑑別函數,可成功的判別83%的塊狀結構魚羣,及93%的規律性結構魚羣,顯示具有良好的判釋效果,證實直接由超音波囘訊特性來區別魚羣的單羣結構是可行的。

關鍵字

核能三廠 漲潮 落潮 溫排水 等溫線

並列摘要


When applying acoustic system to estimate the fish abundance,the result isthat bias are usually subject to internal and external factors that affect theschooling structures of the fish school.Therefore,for improving the accuracyof acoustic estimation,it is necessary to develop specific acoustic criteria todistinguish the signals of different schooling structures.This experiment wasconducted by taking paired stereo photographs of 44 schools of E.japonica;for analyzing the relative positions of individual fish.The Taylor(1961),Clark and Evans(1954)methods were used to distinguish one by one,to determinewhether the school belonged to the regular or clumpy spacing structure.Then for each continuous echo signal from the photographed schools,sevenfundamental explanatory variables were extracted by signal processing.Finally,by employing the stepwise discriminant analysis and discriminant analysis,thebest discriminators of different schooling structures were extracted,from which,a model that could directly indentify the schooling structure was establish toexplore more accurate estimation by monitoring the change of schooling structurein the echo signal of E.japonica.The abstract is as follows:(a)From the seven explanatory variables of the echo extracted by usingstepwise discriminant analysis,it was found that the combination of threeindependent variables including echo level of continuous pulse,change in echowidth and number of pulses could be used to identify the spacing structure offish school with the best discriminating result.(b)Regular spacing structure schools were characterized by fewed peaksin continuous echo signals,stability in both echo signal and phase;while clumpyspacing structure was characterized by more peaks and greater changes in echolength,echo level and phase.(c)From the above three discriminators of echo with the best distinguishingresults,the discriminant function was established by applying discriminantanalysis to identify successfully 83% of clumpy spacing structure and 93% ofregular spacing structure.Judging from this fact,indicates that the discriminantfunction can best identify the different structures of fish school,andit also proves that using echo signal processing directly for distinguishingschooling structure is feasible.

延伸閱讀