ARPEGE Cloud Cover Forecast Post-Processing with Convolutional Neural Network

View More View Less
  • 1 Institut de Recherche Technologique Saint-Exupéry, Toulouse, France
  • 2 Météo-France, Direction des Opérations pour la Production, 42 avenue Gaspard Coriolis, 31057 Toulouse cedex 07, France and CNRM/GAME, Météo-France/CNRS URA 1357, Toulouse, France
  • 3 IRIT, Université Paul Sabatier, Toulouse, France
  • 4 Institut de Recherche Technologique Saint-Exupéry, Toulouse, France
  • 5 Météo-France, Direction des Opérations pour la Production, 42 avenue Gaspard Coriolis, 31057 Toulouse cedex 07, France and CNRM/GAME, Météo-France/CNRS URA 1357, Toulouse, France
  • 6 Institut de Recherche Technologique Saint-Exupéry, Toulouse, France
  • 7 CERFACS, Toulouse, France
© Get Permissions
Restricted access

Abstract

Cloud cover provides crucial information for many applications such as planning land observation missions from space. It remains however a challenging variable to forecast, and Numerical Weather Prediction (NWP) models suffer from significant biases, hence justifying the use of statistical post-processing techniques. In this study, ARPEGE (Météo-France global NWP) cloud cover is post-processed using a convolutional neural network (CNN). CNN is the most popular machine learning tool to deal with images. In our case, CNN allows the integration of spatial information contained in NWP outputs. We use a gridded cloud cover product derived from satellite observations over Europe as ground truth, and predictors are spatial fields of various variables produced by ARPEGE at the corresponding lead time. We show that a simple U-Net architecture (a particular type of CNN) produces significant improvements over Europe. Moreover, the U-Net outclasses more traditional machine learning methods used operationally such as a random forest and a logistic quantile regression. When using a large number of predictors, a first step toward interpretation is to produce a ranking of predictors by importance. Traditional methods of ranking (permutation importance, sequential selection, . . . ) need important computational resources. We introduced a weighting predictor layer prior to the traditional U-Net architecture in order to produce such a ranking. The small number of additional weights to train (the same as the number of predictors) does not impact the computational time, representing a huge advantage compared to traditional methods.

Corresponding author: Florian Dupuy, florian.dupuy@meteo.com

Abstract

Cloud cover provides crucial information for many applications such as planning land observation missions from space. It remains however a challenging variable to forecast, and Numerical Weather Prediction (NWP) models suffer from significant biases, hence justifying the use of statistical post-processing techniques. In this study, ARPEGE (Météo-France global NWP) cloud cover is post-processed using a convolutional neural network (CNN). CNN is the most popular machine learning tool to deal with images. In our case, CNN allows the integration of spatial information contained in NWP outputs. We use a gridded cloud cover product derived from satellite observations over Europe as ground truth, and predictors are spatial fields of various variables produced by ARPEGE at the corresponding lead time. We show that a simple U-Net architecture (a particular type of CNN) produces significant improvements over Europe. Moreover, the U-Net outclasses more traditional machine learning methods used operationally such as a random forest and a logistic quantile regression. When using a large number of predictors, a first step toward interpretation is to produce a ranking of predictors by importance. Traditional methods of ranking (permutation importance, sequential selection, . . . ) need important computational resources. We introduced a weighting predictor layer prior to the traditional U-Net architecture in order to produce such a ranking. The small number of additional weights to train (the same as the number of predictors) does not impact the computational time, representing a huge advantage compared to traditional methods.

Corresponding author: Florian Dupuy, florian.dupuy@meteo.com
Save