Simplifying Land Cover-Geoprocessing-Model Migration with a PAMC-LC Containerization Strategy in the Open Web Environment

Abstract

Land cover and its changes over time are significant for better understanding the Earth’s fundamental characteristics and processes, such as global climate change, hydrology, and the carbon cycle. A number of land cover-geoprocessing models have been proposed for land cover-data production with different spatial and temporal resolutions. With the massive growth in land cover data and the increasing demand for efficient model utilization, developing efficient and convenient land cover-geoprocessing models has become a formidable challenge. Although some model-migration methods have been proposed for handling the massive data, the intricacy of land cover-data and -heterogeneity models frequently prevent current strategies from directly meeting demand. In this paper, we propose the PAMC-LC-containerization approach to overcome the difficulties associated with moving existing land cover models in the open web environment. Based on the idea of model migration, we design a standardized model description and hierarchical encapsulation strategy for land cover models, and develop migration and deployment methods. Furthermore, we assess the viability and efficacy of the proposed approach by using coupled workflows for model migration and the introduction of visualization on the Mts-WH dataset and the Google dataset. The experimental results show that the PAMC-LC approach can simplify and streamline the model migration process, with important ramifications for increasing productivity, reusing models, and lowering additional data-transmission costs.