ANR Project T-REX
New TRends in EXtremes,
prediction and validation





    Forecast is a major task of statistics in many domains of application. It often takes the form of a probabilistic forecast where the so-called predictive distribution represents the uncertainty of the future outcome given the information available today. Of particular interest is the  distributional forecast of rare and extreme events, for instance environmental hazards such as flooding or heat waves, that can have major  socio-economic consequences but for which current prediction is often inaccurate and unsatisfactory.

    In meteorology and weather forecast, the current numerical weather prediction (NWP) models are rather skillful for non-extreme weather events but often fail to provide accurate predictions on extreme weather events. One objective is to derive new statistical post-processing methods that are tailored to the output of existing NWP models in order to improve the forecast of extreme weather. Tree-based method such as generalized random forest for extremes will be developed. Data sets required to develop and validate the new proposed methods will be provided by the Centre National de Recherches Météorologiques (CNRM).

    In energy and electricity consumption forecast, balancing production and demand is a major concern and forecasting the demand and especially its peaks is crucial. Electricit\'e de France (EDF) has developed a strong expertise where one key tool is the possibility to combine different models to improve prediction. Aggregation of experts is another main direction of the T-REX project with an emphasis on specialized/sleeping experts that focus on extreme regimes. EDF will provide relevant data sets for electricity consumption forecast.

    Extreme value theory (EVT) provides a theoretical framework for risk assessment and mathematically justified estimation of rare event probabilities. The T-REX project is rooted in EVT and will bridge different research fields to improve probability forecasts of extremes from complex and high dimensional systems.

    The structure of the project is as follows:
        A) Extreme value theory
                A1) Extreme quantile regression and the proportional tail model
                A2) Tree based method for extreme prediction
                A3) Multivariate extreme value theory
        B) Prediction and validation
                B1) Prediction by aggregation of experts
                B2) Validation of prediction
        C) Applications
                C1) Statistical post-processing of numerical weather predictions and extremes
                C2) Online forecast of energy consumption and extremes


The workshop Valpred 3 is organized at the Centre Paul Langevin in Aussois with more than 40 participants !
Romain Pic is hired for a PhD on the statistical post-processing of probabilistic weather forecast  jointly supervised by Clément Dombry (LmB), Philippe Naveau (LSCE) and Maxime Taillardat (CNRM) .
A contributed session on "Prediction and validation for extremes" is organised at the Extreme Value  Analysis Conference EVA 2021.
Romain Pic is hired for a Master internship (6 months) and will work on the prediction of extreme weather forecast.
The T-REX project is officially launched for 4 years !


To acknowledge funding, please mention "The authors acknowledge the support of the French Agence Nationale de la Recherche (ANR) under reference ANR-20-CE40-0025-01 (T-REX project)". The following research has been partly founded by the ANR project.
  1. M. Taillardat. Skewed and Mixture of Gaussian Distributions for Ensemble Postprocessing. Atmosphere, 12(8), 966, 2021.journal
  2. P. Rivoire, O. Martius and P. Naveau. A Comparison of Moderate and Extreme ERA-5 Daily Precipitation With Two Observational Data Sets. Earth and Space Science, 8(4), 2021. journal
  3. J. Bessac and P. Naveau. Forecast score distributions with imperfect observations. Advances in Statistical Climatology, Meteorology and Oceanography, in press, 2021. journal
  4. J. Velthoen, C. Dombry, J.J. Cai and S. Engelke. Gradient Boosting for extreme quantile regression. arXiv:2103.00808
  5. A. Antoniadis, S. Gaucher and Y. Goude. Hierarchical transfer learning with applications for electricity load forecasting. hal-03429702
  6. S. Mentemeier and O. Wintenberger. Asymptotic independence ex machina: Extreme Value Theory for the diagonal SRE model.  Journal of Time Series  Analysis, in press, 2021. journal