Dynamic 3D Visualization of Climate Model Development and Results

Jeremy Walton, Samantha Adams, Wolfgang Hayek, Piotr Florek, Harold Dyson

External link (DOI)
View presentation:2021-10-27T17:00:00ZGMT-0600Change your timezone on the schedule page
2021-10-27T17:00:00Z
Exemplar figure, described by caption below
Four possibilities for temperatures at the Earth’s surface in 2100, determined by our climate model (UKESM1) using four scenarios for future emissions of greenhouse gases such as CO2. These range from one where the world adopts “sustainable development” to a “business as usual” Earth which continues to rely on fossil fuels, resulting in rising emissions of CO2. Warming increases with emissions levels, and is not distributed evenly around the world. Thus, although the “middle of the road” scenario sees a global average temperature rise of 4 degrees by 2100, parts of the Arctic warm by as much as 20 degrees.
Fast forward

Direct link to video on YouTube: https://youtu.be/OcscgKSYk_8

Abstract

Climate models play a significant role in the understanding of climate change, and the effective presentation and interpretation of their results is important for both the scientific community and the general public. In the case of the latter audience-which has become increasingly concerned with the implications of climate change for society-there is a requirement for visualizations which are compelling and engaging. We describe the use of ParaView, a well-established visualization application, to produce images and animations of results from a large set of modeling experiments, and their use in the promulgation of climate research results. Visualization can also make useful contributions to development, particularly for complex large-scale applications such as climate models. We present early results from the construction of a next-generation climate model which has been designed for use on exascale compute platforms, and show how visualization has helped in the development process, particularly with regard to higher model resolutions and novel data representations.