RMWSPy (v 1.1): A Python code for spatial simulation and inversion for environmental applications

• Published in: Environmental modelling & software : with environment data news Vol. 138; p. 104970
• Main Authors: Hörning, Sebastian, Haese, Barbara
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.04.2021; Elsevier Science Ltd
• Subjects: Algorithms; computer software; Environment models; Environmental modeling; Environmental modelling; Environmental science; Fields (mathematics); Geostatistical simulation; Groundwater; Groundwater flow; Precipitation; Programming languages; Python; Rain; Rain gauges; Random mixing; Simulation; Stochastic inversion; Geostatistical simulation; Random mixing; Stochastic inversion; Environmental modelling; Python
• ISSN: 1364-8152; 1873-6726
• DOI: 10.1016/j.envsoft.2021.104970

Spatial simulation and inversion problems are omnipresent in earth and environmental sciences. An open-source Python package (RMWSPy) for conditional spatial random field simulation and inversion based on a generalized implementation of the Random Mixing Whittaker-Shannon (RMWS) algorithm is presented in this paper. The RMWS algorithm has successfully been applied to a variety of environmental modelling problems, ranging from inverse groundwater flow and transport modelling to precipitation simulation incorporating incomplete observations. RMWSPy provides great flexibility due to its variety of linear and non-linear conditioning constraints. The generalized implementation isolates the core algorithm from the user-defined problem statement. In this paper, RMWSPy is introduced using a synthetic inversion example for spatial precipitation estimation which combines rain gauge data and integral rain rates obtained from Commercial Microwave Link data. The required Python scripts are described and the results of one precipitation event are presented and discussed. •Combined geostatistical simulation and inversion algorithm/code for environmental modelling problems.•Generalized implementation for plug-and-play like use.•Flexible with a wide variety of use cases.•Simple Python-based code + example applications.