IMAGe's 2016 Theme of the Year

By Marijke Unger
06/30/2016 - 12:15am

CISL’s Institute for Mathematics Applied to Geosciences (IMAGe) is dedicated to applying mathematical models and conceptual tools to fundamental problems in the geosciences, while serving as a nexus of activity for the mathematical and geophysical communities.

The division brings together computational mathematics, data assimilation research, geophysical statistics, and regional integrated sciences, all under the leadership of Doug Nychka, who also developed one of IMAGe’s flagship efforts, an annual series that takes a thematic approach to advancing solutions to important and complex scientific questions.

Known as Theme of the Year, or TOY for short, the series consists of workshops, conferences, presentations, and training opportunities around a fundamental problem in the geosciences.

“The main idea for TOY was to focus on specific NCAR research topics that seemed particularly well-suited to applied math,” said Nychka.

This year’s theme focuses on extremes in climate sciences, and is being organized by visiting scientist Philippe Naveau, a leading expert in the field. Extreme value theory has long been an active area of mathematical statistics but it is only recently that approaches have been devised to handle the practical problems for large climate data sets. One goal of this TOY is encourage this process of technology transfer between the data science and the work at NCAR. 

Philippe Naveau
Philippe Naveau teaches a course related to the 2016 Theme of the Year (photo by Brian Bevirt).

Naveau began his career as a postdoc at NCAR, after completing his studies at CSU. Originally from France, he returned there after being an assistant professor at CU (Applied Math) to continue his work at France’s Centre National de Recherche Scientifique (CNRS), in the Laboratoire des Sciences du Climat et l’Environnement.

He has maintained close ties to NCAR, CU, and CSU researchers working on climate extremes, and his current year-long tenure with IMAGe is providing the perfect opportunity for collaboration among these groups and other researchers and institutions around the world, while enhancing TOY, IMAGe, and CISL’s programs.

Formally entitled “Extremes in Climate Sciences: a Statistical, Dynamical, and Machine Learning Inquiry,” the 2015-2016 TOY has brought together national and international experts and presenters from CSU, NOAA, CU, and NCAR. The goal of this year’s focus is to improve ways of analyzing data in hopes of developing more accurate tools for modeling the frequency and intensities of extreme events, among other applications.

The TOY series kicked off with monthly presentations starting in October of 2015. Among them were two sessions on extremes as part of the International Detection and Attribution Group (IDAG), in February, and Beyond P-Values: The Statistics, a training program for NCAR staff, both featured in CISL’s summer newsletter.

In addition to events and workshops hosted at NCAR, Doug Nychka and Philippe Naveau also participated in the Banff International Research Station workshop on “Uncertainty Modeling in the Analysis of Weather, Climate, and Hydrological Extremes” in June, exposing a new audience to NCAR science.

While breakthroughs in statistical analysis can be hard to quantify, the TOY has contributed to a distinct refocusing of this field. Classical extreme value theory has usually remained distinct from analysis of the more frequent values, namely, those around the mean.

“What has emerged from the TOY is a plan for data analysis that can seamlessly move from characterizing the center of a distribution to that of the tails of the bell curve,” said Nychka. “Not having to choose between standard versus extreme value methods has the advantage of producing an intermediate model that more closely resembles natural processes.”

A full listing of the TOY events and presentations is available here