ADS Highlights Event 2022

ADS will host its annual Highlights Event in-person Thursday 8th of December 2022! The event will also feature highlights from our academic and industry partners and the winners of the third ADS Thesis Awards. 

Programme 

14:00   14:30   Welcome
14:30   14:40   Introduction
14:40   15:15   Keynote
15:15   15:25   Q&A
15:25   15:50   Break
15:50   16:15   ADS Thesis Awards
16:15   16:45   Partner Updates
16:45   16:55   Closing remarks
16:55   18:30   Borrel & Poster Presentations

Dim Coumou (KNMI) will give a keynote on:
Entering the Era of Big Data in Climate Science:
Our climate is changing rapidly and people and societies around the globe feel that via changes in extreme weather events like heat waves, droughts or extreme rainfall. Last summer dramatically illustrated this, with record-drought in Europe, heat waves all over the northern hemisphere, and extreme rain and massive flooding in Pakistan. Indeed heat waves and extreme rainfall events have strongly increased worldwide over the last century. These rapid changes provide new challenges both for society and climate scientists.

In his talk, he will present recent research showing how machine learning and Big Data can help addressing some of the more pressing challenges. The total volume of climate data, from observations like satellites but also from climate models, is now estimated at 100s of petabytes. In addition we have large impact data sets like crop data from satellites or critical infrastructure derived from open street map. This data provides enormous opportunities for climate scientists, but the challenge is to create actual new knowledge from this deluge of data.

Applying artificial intelligence (AI) methods in climate sciences is currently a rapidly developing field, creating high hopes that such tools can push understanding and predictability of the climate system. Machine learning methods are particularly well suited to improve predictability of extremes at relatively long lead times, i.e. beyond two weeks. Forecasting weather beyond two weeks is notoriously difficult, and current operational systems based on weather or climate models, have little or no predictive power at these timescales. New research from our group shows how AI can identify low frequency processes in the climate system that can be used to forecast extremes several weeks to months ahead. For example, slowly changing patterns of warm surface waters in the Pacific ocean can give predictability of eastern United States summer temperatures. Moreover, these same ocean patterns can also be used to directly forecast crop failures in this region. I my talk, I will discuss this as well as other examples to illustrate how advanced machine learning can give us improved predictability of extremes as well as improved physical understanding of the underlying processes.

About KNMI
The Royal Netherlands Meteorological Institute (KNMI) is the Dutch national weather service. Primary tasks of KNMI are weather forecasting and monitoring of weather, climate, air quality and seismic activity. KNMI is also the national research and information centre for meteorology, climate, air quality, and seismology. KNMI provides reliable and consistent measurements, data and forecasts that form the basis of important decisions that keep the Netherlands safe. From a code red for road traffic to the climate scenarios for the Delta Program involving billions of euros. For a safe Netherlands that is prepared for the influence of weather, climate and earthquakes.

This event is on an invitation-only basis. Contact info@amsterdamdatascience.nl if you are an ADS Partner and haven’t received an invitation and would like to attend!