A Study On Encapsulation And Integration Of Geographic Model Considering Data And Function Classification

In recent years, with the deep exploration of geoscience research, the quantity and quality of geographic model resources have been greatly improved. But due to the differences in the semantics and modeling environment of geographic models, the reuse of them had been greatly hindered, which results in the waste of geographic model resources. Because of the lack of a consistent data description and the unified standards of geographic model integration, traditional geographic models’encapsulation and integration methods can not meet the requirement of geographic analysis. The implementation and application of model integration are mostly face to developers rather than users, so there has been a lack of decision-making guidance, especially for users without professional background. Conclusion demonstrated that it’s an urgent need to build a geographical model integrated method which has close unitarity and effective linkage of data, model and function.The object-oriented (00) geography model expression method is an effective technical measure to encapsulation and integration of geographical models. The geography data and geographical models were concluded and integrated at first. Then we build the classification system of geographic data as same as the classification and coding system of geographical model base on the existing metadata standards. The00idea was adopted during the design of unified geographic data interface, on the purpose of realizing the universal access and management of geographical data. On the other hand, the abstract base classes of geographical models should be constructed to realize the geography model initialized, data configuration and operation parameters configuration in the form of inheritance class of geographical model, which can provide model libraries support of geographical model encapsulations and integrations. With the theoretical guidance to the rules of geographic models discovery, based on the geographic data characteristics and application requirements, the geographic model chain can be built and we designed the overall framework of geographic model integration to achieve the data-oriented and function-oriented geographical model encapsulation, integration and model template building. Finally, we designed the XML based storage and retrieval mechanism of geographic model template, by this way, the reusability of model template was improved, which can provide effective support to follow-upgeographic data analysis.Take the before-mentioned encapsulation and integration methods of geographic models as the theoretical guidance, we implemented the prototype system in C++language. The meteorological data in2005of China was used to demonstrate our methods.Through the characteristic analysis of meteorological data, combined with practical application,we constructed a model chain for the exploratory geostatistics analysis of meteorological data, and saved it as a model template. After that, the precipitation data in China is applied to this modeltemplate, based on which we realized the model integration and model computation of the average temperature data and precipitation. Meanwhile, a discussion of the statistical analysis and spatial distribution trend of average temperature and precipitation was made.This case clearly demonstrates that the proposed encapsulation and integration methods of geographical model whichgive consideration to both data and function has great generality, and it can provides effective quotation and reference for model reuse and integration.