基于广义可加模型和案例推理的东南太平洋智利竹筴鱼中心渔场预报

Prediction of central fishing ground of Chilean jack mackerel(Trachurus murphyi) in the Southeast Pacific Ocean based on generalized additive model and case-based reason

根据2001～2008年我国渔船在东南太平洋海域的智利竹筴鱼生产统计数据以及同期卫星遥感数据反演的海表温度、叶绿素浓度、海表温度梯度等数据,利用广义可加模型定量分析了智利外海竹筴鱼资源分布同环境因子的关系。根据GAM模型的研究结果,确定海表温度作为竹筴鱼中心渔场预报指标。利用案例推理方法,通过三级相似检索对智利竹筴鱼中心渔场进行预报。试验性预报实例的结果与渔船实际作业情况比较表明,预报精度达到68%,能一定程度上反映竹筴鱼资源的分布。

Based on the Chilean jack mackerel（Trachurus murphyi） fishing data from Chinese vessel in the Southeast Paciffic Ocean during 2001~2008,as well as the sea surface temperature（SST）,chlorophyll concentration,and temperature gradient（TGR） derived from the satellite remote sensing data,the relationship between the distribution of jack mackerel offshore waters of Chile and environmental factors were analyzed quantitatively using general additive model（GAM）.According to the result of GAM,SST was confirmed to be the index used to predict central fishing ground.Using case-based reasoning method,through three class similar searching,the central fishing ground of Chilean jack mackerel was predicted.As an example of the results,compared the experimental central fishing ground forecasted with fishing catches by the vessels,the result showed that the precision was 68%,which could reflect the distribution of Chilean jack mackerel in a certain extent.