Study On Parallelizing And Visualization Of Offshore Ecological Environment Evaluation System

Due to its large and complex process, ocean numerical model has an extremely large scale of computing, and as a result, its calculation contains huge amounts of data. Using parallel computing can significantly enhance the computing efficiency, in order to complete the calculation tasks in the expect time. In the process of applying parallel technology, the optimization of parallel programs is an integral part to improve the overall performance of the application, and the I/O optimization is one of the key points.On the basis of completing application efficiency optimization, using data visualization and Mashup technology could effectively help to find out the appearance and association hidden in the data of calculated results, as well as the relationships between numerical data and reality geographic information, it can provide strong help for us to understand and analyze the data and find out the regularities.In this paper, MPI-based parallel technology was used on the model ofcoupled physical-ecosystem in the coastal ecological environment evaluation system of Jiao Zhou Bay demonstration zone, and then further optimize was applied in the parallel application, after that the data visualization technology was used in this application and the visual result was mashuped with the web map, main tasks are as follows:The original serial mode has been parallelized and the effectiveness enhancing was tested and evaluated, mainly implemented the parallelization of the coupled physical-ecosystem model in Jiao Zhou Bay demonstration zone, the main idea was based on dividing the whole region to several sub regions, each sub region was calculated by a different computing node, thereupon then the efficiency of the whole system will be improved;While the mode of calculation module has been parallel optimized, the I/O modules were still run asa traditional serial way, to a certain extent, it willreducethe performance of the application. Thus MPI-2 standards based parallel I/O algorithms was used in the I/O modules, and comparison of parallel I/O with the traditional serial I/O was made;After the efficiency upgrade of the application, HTML5-based visualization technology was used to visualize the calculation results of ocean model, then with the Mashup technology the calculations of ocean model and realistic geographic information was combined. Through the AJAX engine based technology,interactive data extraction and images display was used make the appearances and correlation hidden in the data as well as the relationships between data and reality became visualized. It provided a powerful method for analyzing data and finding out rules in the ocean data.Results showed that by using parallel computing and parallel I/O technology in the coupled physical-ecosystem model of Jiao Zhou Bay demonstration zone, the operating speed was greatly enhanced; Mashup-based visualization technology provides a powerful help to the data mining and effectively information exchange among researchers, also scientific data could be served for policy makers and ordinary users in a better way.