Graduate Students Contribute a Novel Open-Source Forecasting Verification Tool to the Pangeo Community
Verification of ocean pH forecasts using climpred. “Lead years” refer to how many years in advance the model is predicting ocean pH. The top two rows show the accuracy of the model forecast system and a statistical forecast at predicting observations, respectively. Red denotes a more accurate forecast in the top two rows, and brown denotes an improvement in forecast quality by using a dynamical forecasting system rather than a statistical forecast-based solely on the observations. Credit: Riley Brady, graduate student at University of Colorado at Boulder, and Aaron Spring, graduate student at the Max Planck Institute in Hamburg, Germany.
December 18, 2020
By: Kimberly Mann Bruch
Riley Brady, graduate student at University of Colorado at Boulder, and Aaron Spring, graduate student at the Max Planck Institute in Hamburg, Germany, recently collaborated to create an open-source python-based Pangeo library that allows worldwide researchers to easily analyze forecasts using weather and climate models. Coined “climpred” (climate prediction), the new tool has been developed in conjunction with EarthCube’s Pangeo community so that geoscientists can post-process, analyze, and visualize the results of climate prediction systems.
“Some of these models are more than 500 terabytes – that’s 500,000 gigabytes! – and that is a lot of data to analyze,” said Brady, who is a Department of Energy computational science fellow. “Our climpred tool lets geoscientists evaluate how well the forecasts and hindcasts do when predicting the observed world.”
Specifically, climpred was created so that geoscientists can quantify the accuracy of various forecasting and hindcasting models versus actual observations. Brady and Spring have even used climpred to analyze predictions from ocean chemistry models, which is very complex yet important. As more researchers have started focusing studies on ocean acidification, climpred provides a way to better understand the past and future of the ocean’s acidity.
“With Pangeo and through the Python language, we have collaborated with many scientists to develop this software that helps us to analyze massive forecasting datasets – applied to important phenomena like ocean acidification – more easily, quickly, and accurately,” explained Brady. “We have already observed coastal environments under stress due to ocean acidification and our hope is that climpred will be utilized by the Pangeo community and beyond to help understand and predict these complicated environmental issues.”
Climpred recently joined the Pangeo community, which is a project funded by the National Science Foundation EarthCube Award 2026932.
EarthCube is a community-driven activity sponsored by the National Science Foundation to transform research in the academic geosciences community. EarthCube aims to create a well-connected environment to share data and knowledge in an open, transparent, and inclusive manner, thus accelerating our ability to better understand and predict the Earth’s systems. EarthCube membership is free and open to anyone in the Geosciences, as well as those building platforms to serve the Earth Sciences. The EarthCube Office is led by the San Diego Supercomputer Center (SDSC) on the UC San Diego campus.
Kimberly Mann Bruch, San Diego Supercomputer Center Communications, email@example.com
Lynne Schreiber, San Diego Supercomputer Center EarthCube Office, firstname.lastname@example.org
San Diego Supercomputer Center: https://www.sdsc.edu/
UC San Diego: https://ucsd.edu/
National Science Foundation: https://www.nsf.gov/