An environmental modelling framework based on asynchronous many-tasks: Scalability and usability. (May 2021)

• Record Type: Journal Article
• Authors: de Jong, Kor; Panja, Debabrata; van Kreveld, Marc; Karssenberg, Derek
• Is Part Of: Environmental modelling & software. Volume 139(2021)
• Publication Date: 2021-05
• Subjects: Environmental modelling framework -- Map algebra -- High-performance computing -- Asynchronous many-tasks -- HPX -- LUE; Environmental monitoring -- Computer programs -- Periodicals; Ecology -- Computer simulation -- Periodicals; Digital computer simulation -- Periodicals; Computer software -- Periodicals; Environmental Monitoring -- Periodicals; Computer Simulation -- Periodicals; Environnement -- Surveillance -- Logiciels -- Périodiques; Écologie -- Simulation, Méthodes de -- Périodiques; Simulation par ordinateur -- Périodiques; Logiciels -- Périodiques; Computer software; Digital computer simulation; Ecology -- Computer simulation; Environmental monitoring -- Computer programs; Periodicals; Electronic journals; 363.70015118
• DOI: 10.1016/j.envsoft.2021.104998
• Languages: English
• ISSNs: 1364-8152

Abstract

Abstract: Environmental modelling frameworks allow domain experts, rather than software developers, to implement and run numerical simulation models in earth and environmental sciences. Because of the need to use more detailed process representations or larger datasets as input to models, it may become infeasible to perform modelling studies, due to the increased amount of time it takes for models to calculate results. The objective of this study is to evaluate the asynchronous many-task approach in the implementation of a prototype scalable modelling framework. We evaluate the scalability of local, focal, and zonal map algebra operations, and an example model in which these operations are combined. Our results show that the capacity of the operations and the example model to use additional hardware, like nodes in a computer cluster, is good. With our freely available prototype framework, models can be executed faster and modelling studies processing considerably more data can be performed. Highlights: A free to use environmental modelling framework with map algebra operations. Models scale well over CPU cores, NUMA nodes, and cluster nodes. Model developers do not need to know about writing parallel algorithms. Framework developers can add modelling operations that work well in combination.