基于GM(1,1)模型的贫信息渔业数据CPUE标准化研究
CPUE standardization for poor information fishery data based on GM (1,1) model
-
摘要: 单位捕捞努力量渔获量(catch per unit of effort,CPUE)标准化是渔业资源评估和管理中基础性工作.为了对信息量少、没有时空和环境变量的渔业监测数据进行CPUE标准化,本研究采用灰色系统方法,结合2010—2014年江苏省4种张网类调查生产数据,构建灰色GM(1,1) CPUE标准化模型,采用残差、关联度和后验差等3种检验方法评价CPUE标准化优良度,为渔业监测调查数据CPUE标准化提供一种新的途径.结果表明,所有模型的残差都可以接受;在关联度检验下,所有灰色GM(1,1)模型的关联度都大于0.6,建模结果为满意;后验差结果显示所有灰色GM(1,1)的预测精度为合格.上述结果表明采用灰色GM(1,1)模型对信息量少、没有空间和环境信息的渔获数据进行CPUE标准化是可行的,可以为渔业管理部门提供决策支持.Abstract: Catch per unit of effort (CPUE) standardization is a basic work in stock assessment and management. In order to standardize poor information fishery data, a grey GM (1,1) CPUE standardization model was constructed based on the grey system method and four sets of research data of Jiangsu province from 2010 to 2014. The goodness of model was evaluated by the residual and grey correlation between the observed and the corresponding predicted values and aftertest residue checking. The results showed that all the models were acceptable under the residual test. All the grey relational values were larger than 0.6 after the grey relational checks. The aftertest residue checks displayed that all models were acceptable and that the traditional GM (1,1) model is feasible to standardize the poor information fishery data. thus, the results can be used to support the fisheries management.