Global warming is one of the most important issues facing the world over the next century. Its impacts will reverberate through the climate and the biome with ever-increasing effect. In the human sphere, it will transform politics and the global economy as the world we live in changes.
Physical atmospheric simulations do an excellent job of making precise, localized predictions about the weather over the span of a few days. Those same simulations can be used on longer time scales to make integrative predictions about things like accumulated rainfall, temperature changes, average cloud cover, and sea level rise over the span of years and decades, assuming the nature of the atmosphere isn’t changing.
But it is changing. Including climate change signals requires integrating the best scientific models that characterize that change—and how the rate of change itself might be changing over time—in to those physical simulations. Historical data is then used to verify that the enhanced atmospheric simulations are in fact describing the observed reality of the change. All of this requires massive scale, both in terms of datasets consumed and produced and the amount of computing power. Using cloud infrastructure, modern orchestration, and containerization allows these massive workloads to be run on public clouds.
Join Ari Gesher for a talk exploring the data science, traditional science, DevOps, and data engineering that goes into accurately predicting our changed climate future.
Ari Gesher is the founding director of software engineering at Kairos Aerospace, a startup building and operating the next-generation of airborne and spaceborne sensors for monitoring oil and gas infrastructure. Ari also serves as consulting architect for Jupiter, a company productizing high-quality datasets that describe the long-term effects of climate change. Previously, he was a very early engineer at Palantir Technologies and later served as Palantir’s engineering ambassador to the tech community at large; before Palantir, he was the maintainer of the SourceForge.net open source archive. Ari is the coauthor of The Architecture of Privacy, which explains how to responsibly hold data about people while preserving their privacy to the greatest extent possible. Ari is a frequent speaker on various topics, including the need for modern, high-leverage engineers to work on substantive problems, human-computer symbiosis as system design aesthetic, the limits of automated decision making, and privacy architectures for a world where everything is recorded.
©2018, O'Reilly Media, Inc. • (800) 889-8969 or (707) 827-7019 • Monday-Friday 7:30am-5pm PT • All trademarks and registered trademarks appearing on oreilly.com are the property of their respective owners. • firstname.lastname@example.org