极地海冰-海洋参数遥感反演模型分布式共享研究

本文以SOA开放式架构与OGC标准规范,提出了极地海冰-海洋参数遥感反演模型分布式共享服务体系。服务体系以"模型服务"为核心,探讨了模型服务接口和模型服务的互操作问题。为了简化极地海冰-海洋参数遥感反演模型的分布式共享过程,提出了极地海冰-海洋参数遥感反演模型共享服务平台的概念。共享服务平台处于模型与模型应用客户端之间,可以实现两者之间的数据转化和功能协同,以及实现模型算法与其他功能的分离,使模型开发者可以专注于模型算法的设计和实现。最后,以海冰密集度遥感反演模型和冰间湖识别模型为例,实现了极地海冰-海洋参数遥感反演模型分布式共享方法。

Research on Distributed Sharing of Polar Sea Ice-Ocean Parameters Remote Sensing Inversion Models

Traditional single computing environment cannot meet the needs of geographic model sharing, because of its limitations on storage, computing resources and program transfer. Distributed geospatial model sharing could avoid those limitations, so distributed sharing architecture of remote sensing inversion models for polar sea ice-ocean parameters is brought forward based on SOA construction and OGC specifications, which can provide the overall framework and the top-level guidance for studying the key technologies of polar sea ice-ocean parameters remote sensing inversion model service composition and constructing specific composition applications. The distributed sharing architecture focuses on the model services. Detail discussion is carried out on model service interface and interoperation problems related to model services. The polar sea ice-ocean param-eters remote sensing inversion model sharing services platform is designed and developed to help implementing polar sea ice-ocean parameters remote sensing inversion model sharing. In this paper, we analyzed the design guidelines of polar sea ice-ocean parameters remote sensing inversion model sharing service platform, and further studied the key technologies involved in the polar sea ice-ocean parameters remote sensing inversion model sharing service platform. The sharing platform is the connector of model and the clients, and can realize the data conversion and function collaborative. With the help of the sharing platform, model developer could only focus on model algorithm, and the sharing platform will take care of building model service, and interacting with model clients. Several models are adapted, including sea ice concentration remote sensing inversion model and polynya morphologic remote sensing inversion model, to demonstrate the advantages of distributed sharing architecture of polar sea ice-ocean parameters remote sensing inversion models.