Climate Change: 3 tech talks

=AGENDA=

5:00 – 5:10 Introduction

5:10 – 5:40 Anna Kwa – Improving climate models with machine learning

Abstract: Global climate models are in good agreement in predicting temperature increases over the coming decades in response to greenhouse gas emissions. However, there is less consensus around how precipitation patterns and extrema will change in the future. Climate models are simulated at low grid scale resolutions (~100 km) due to long time ranges and limits on computational resources, but many precipitation-associated atmospheric processes are described on much smaller scales (e.g. storm clouds are not 100 km in size). Current methods of approximating these processes at low resolutions are a major source of uncertainty in long-term rainfall predictions. Our team is working on using ML to reduce this source of error in climate models. This problem is particularly challenging because of the continuous feedback loop between the trained ML model and the global climate model it is used in. If the ML overcorrects and moves the atmospheric state too far away from a physical state, the climate model will quickly become unstable.

5:40 – 6:10 Skylar Olsen – Bringing the risk home to drive change

Increasing risks of fire, flood, drought, temperature change, and precipitation will affect property values and transform the mortgage and insurance markets. Millions of dollars of research is being done every year, but the average person isn’t able to access this information to understand their specific risk. Scientific reports can be found but are difficult to understand and interpret. Come join Dr. Skylar Olsen to talk about the massive climate risk database built by ClimateCheck data scientists and the ways she and her team use data science and analytics to bring the risk home to the individual.

6:10 – 6:30 Katinka Bellomo – Leveraging data science to reduce uncertainty in predictions of future climate change

The Atlantic Meridional Overturning Circulation, for short ‘AMOC’, is the largest, global-scale ocean circulation. In response to global warming, the strength of the AMOC is decreasing. However, climate models used to predict future climate change exhibit a large inter-model spread in the possible decline rates of the AMOC, which drives large uncertainty in the prediction of climate change impacts. In this talk, I will show what could happen in a future with a small AMOC decline versus a future with a large AMOC decline. I will also discuss how data science can be leveraged to reduce the uncertainty in the prediction of the AMOC response to global warming.

6:30 – 7:00 virtual networking

=BIOS=
Anna Kwa, PhD: Anna completed a PhD in physics from UC Irvine in 2017, where she researched astro-particle physics. Afterwards she moved to Seattle to work in data science. She is currently working at Vulcan on a team dedicated to improving climate models.

Skylar Olsen, PhD: Skylar is Principal Economist at ClimateCheck, a new startup dedicated to providing easy to understand and detailed information about Climate Change induced risks to properties, lifestyles, markets, and communities. Previously Senior Principal Economist and Director of Economic Research at Zillow, she was a foundational member of the economic research program at Zillow and directed many of its data sharing initiatives. She holds a PhD in economics with a specialization in econometrics and environmental & resource economics from the University of Washington.

Katinka Bellomo, PhD: Katinka is a climate scientist at the Institute of Atmospheric Sciences and Climate in Turin, Italy. She uses networks of ground and satellite observations, global climate models, statistics, and data science to investigate mechanisms of large-scale climate variability and change. You can follow her research and outreach activities at https://theclimatescientist.com and on Twitter (@katinka_bellomo).

=CODE OF CONDUCT=
https://github.com/WiMLDS/starter-kit/wiki/Code-of-conduct