Seminar: How Can We Detect Statistically Robust Climate Trends?

Sign in
September 28, 2023
September 28th, 2023
CRM Auditorium / Zoom



Climate policies are based in part on quantitative projections of climate trends; for instance, the Montreal protocol was based on projections of critical ozone loss. One key aspect of such projections is the concept of time of emergence, that is, the year (or decade) beyond which an observed trend is expected to emerge from natural variability. I will address standing issues with methods currently in use to estimate this time of emergence:

  • Issue 1: the lack of proper formalism to quantify the statistical confidence placed in the emergence, which is a key ingredient for the development of climate policy;
  • Issue 2: the inability to account for observational limitations (varying sampling, smoothing, gaps, etc), which is problematic when analyzing historical trends.

I will illustrate my talk with discussions about the stratospheric ozone layer and the mean position of the subtropical jets. This seminar is part of the International Programme for Research in Groups in collaboration with Dr. Jezabel Curbelo (UPC).


Louis Rivoire

Harvard University

I am primarily interested in the role of the upper troposphere and lower stratosphere in the climate. I also work to understand how sampling biases and other observational limitations affect our understanding of the climate.

My work is driven by applications in atmospheric dynamics and composition, air quality, hurricane risk, and the development of new satellites.

I graduated with a Master’s in Ocean/Atmosphere/Climate/Remote Sensing Sciences from Ecole Normale Supérieure/Sorbonne University in 2015. My PhD work was done at Colorado State University and investigated cold anomalies above hurricanes. I am currently a postdoctoral fellow at Harvard University, and will soon expand my research as a postdoctoral fellow at MIT.


Professor Rivoire’s visit is part of the CRM International Programme for Research in Groups (IP4RG).