Bayes Fishing Ground Forecast Model Optimization And WebGIS Implementation

The South China Sea(SCS), an significant fishing ground in China, is vast and rich in resources. Fisheries resources in inshore of the SCS depleted due to overfishing while south-central open sea still has greater development potential. The yellowfin tuna(Thunnus albacares) is one of main large tuna fishes, accounting for a large proportion of the gross catches of the large tuna fishes in the SCS. Fishing condition analysis and forecast is an important content of fishery science research. The use of satellite remote sensing(RS) data and the yellowfin tuna fishery catches data to research the relationship between the temporal-spatial distributions of the fishing grounds and sea surface environment and to forecast the fishing grounds have significant scientific values on the exploitation and management of fishing grounds in the open SCS.In this paper, the SCS and adjacent waters(0°-25°N,105°E-125°E) was taken as the research area. The collected ocean environmental RS dataset and the yellowfin tuna fishery data were adopted to analyze their temporal-spatial distributions and relationship characteristic. In order to explore the optimal fishing ground forecast model, there are eight optional models based on Bayes classifier were built. Then based on the optimal model and combined the network technologies with Geographic Information System technologies, a tuna fishing ground forecasting information service Web GIS application platform for the SCS was developed and implemented. The main research contents, method and conclusions of the study can be summarized as follows:(1) Based on the monthly averaged SST data products provided from the Climate Prediction Center of the National Oceanic and Atmospheric Administration(CPC/NOAA) and the yellowfin tuna longline fishing catch and effort data of the South China Sea and its adjacent waters published by the Western and Central Pacific Fisheries Commission(WCPFC), thirty years(1982-2011) of data were extracted out from them and spatial overlay maps of averaged SST and catch per unit effort(CPUE) in each month were plotted to analyze the potential relationship between yellowfin tuna fishing grounds distribution and SST in the South China Sea and its adjacent waters, and then the method of frequency analysis and empirical cumulative distribution function(ECDF) were adopted in order to find out the most suitable SST range for yellowfin tunas inhabiting. The results showed that the yellowfin tuna fishing datasets occurred only in areas where SST ranged from 16℃ to 30℃. In spring and summer(from March to August), most of higher CPUEs occurred in areas between 10°N and 20°N, and CPUE tended to decrease outside of these areas. In autumn and winter(from September to February), the fishing grounds extended to more southern areas. A scatter plot remarkably reveals a negatively skewed distribution of SST by CPUE. High CPUEs mainly occur in areas where SST ranges from 26 ℃ to 30℃ and the highest CPUE appears near 29℃;the distribution of CPUEs tends to be scattered in SST ranged from 22℃ to 26℃ in which quite a number of high CPUEs also occur; below 22℃, the CPUEs mostly belong to low CPUE and null CPUE; respectively, the SST range of null CPUEs, low CPUEs and high CPUEs are 26.7℃(±3.2℃), 27.8℃(±2.1℃) and 28.4℃(±1.5℃), which indicates that the high CPUEs distribution in SST intervals is more concentrated than the low CPUEs as well as the null CPUEs. The SST range from 26.9℃ to 29.4℃can be regarded as an indication of the most suitable habitat for the yellowfin tunas in the South China Sea and its adjacent waters.(2) In order to provide the reliable model basis and highlight the scientific of the fishing ground forecast. The issue how different choice of environment factor and fishing zones classification strategies impact on classification results and precision was explored to optimize the Bayes fishing ground forecast model. Using the yellowfin tuna longline fishing catch data in the open SCS provided by WCPFC, the optimum interpolation sea surface temperature(OISST) from CPC/NOAA and multi-satellites altimetric monthly averaged product sea surface height(SSH) released by OCC/CNES, eight plans were made based on Bayes classifier according to different strategies on the choice of environment factor and classification of fishing zones to classify the YFT fishing ground in the open SCS, compared the classification results with the actual ones for validation and analyzed how different plans impact on classification results and precision. The results of validation show that the precision of the eight plans are 71.4%, 75%, 70.8%, 74.4%, 66.7%, 68.5%, 57.7% and 63.7% in sequence, the first to sixth among them above 65% would meet the practical application needs basically. The plans which use SST and SSH simultaneously as environmental factor have higher precision than which only use single SST environmental factor, and considering the addition of SSH can improve the model precision to a certain extent. The plans which use CPUE’s mean ± standard deviation as threshold have higher precision than which use CPUE’s 33.3% and 66.7% quantiles as the threshold.(3) A tuna fishing ground forecasting information service Web GIS application platform was developed and implemented. The designing thought for the open SCS with its actual requirement was put forward. The primary modules and their functions were described and the key technologies involved were introduced. Finally, the Web GIS application adopts three-tier architecture, using Java as the development language coded in Eclipse, Geoserver as the map server, and Open Layers 3 library to realize the Web GIS client side. The Web GIS application was designed to realize tuna fishing ground quickly forecasting in the open SCS in order to provide professional guidance for the fishery producers and the managers in real time and ensure the sustainable development of the marine fishery. Compared with traditional C/S service systems, the system we developed can be accessed on any Web node by users with no platform dependency and it features high flexibility and favorable expansibility.The innovative work of this paper is as followed:(1) Discussed detailedly the temporal-spatial distributions and relationship characteristic between SST and yellowfin tuna fishing ground CPUE in the SCS and adjacent waters, and obtained the most suitable SST range for yellowfin tunas inhabiting.(2) Made and implemented eight optional plan based on Bayes classifier, explored the issue how different choice of environment factor and fishing zones classification strategies impact on classification results and precision, and provides a new model optimization idea for solving the scientific issues of fishing ground forecast.(3) Developed and implemented a tuna fishing ground forecasting information service Web GIS application platform for the SCS and adjacent waters. The three-tier architecture makes the system stable and easy to maintain, embodies the processing style of modern information technology and programming thinking of “high cohesion and weak coupling”, equips the advantages of high flexibility and favorable expansibility, and provides a sound solution for solving the effectiveness of fishing ground forecast.