Mimi.jl: A Computational Platform for Integrated Assessment Modeling
This paper presents the Mimi platform, a harmonized, user-friendly computational platform for building and running integrated assessment models, enabling cross-institutional collaborations and non-academic use cases.
Abstract
Integrated assessment models that couple representations of human and natural systems play a central role in climate economics research and climate policy analysis, but their interdisciplinary scope and increasing complexity demand more advanced computational platforms. The fragmented ecosystem of tools and programming languages used for these models slows research, impedes transparent policy analysis, complicates cross-institutional collaboration, and creates barriers for newcomers to programming. This paper presents Mimi, an open source software platform designed to address these challenges. We present Mimi’s key design features and ways they align with the goals of computational performance, ease of use for computer science novices, support for modular workflows and disparate teams, and facilitation of transparent, reproducible analysis. We highlight Mimi’s adoption in academic and public-sector contexts. Mimi is an example of how careful design of new computational tools can support novel research and its application to policy.
Keywords: Climate change policy, open source software, integrated assessment modeling, domain specific language, Julia programming language, reproducible research
