Preprints
https://doi.org/10.5194/npg-2021-20
https://doi.org/10.5194/npg-2021-20

  17 May 2021

17 May 2021

Review status: this preprint is currently under review for the journal NPG.

Using neural networks to improve simulations in the gray zone

Raphael Kriegmair1, Yvonne Ruckstuhl1, Stephan Rasp2, and George Craig1 Raphael Kriegmair et al.
  • 1Meteorological Institute Munich, Ludwig-Maximilians-Universität München, Germany
  • 2ClimateAi, Inc.

Abstract. Machine learning represents a potential method to cope with the gray zone problem of representing motions in dynamical systems on scales comparable to the model resolution. Here we explore the possibility of using a neural network to directly learn the error caused by unresolved scales. We use a modified shallow water model which includes highly nonlinear processes mimicking atmospheric convection. To create the training dataset we run the model in a high and a low-resolution setup and compare the difference after one low resolution time step starting from the same initial conditions, thereby obtaining an exact target. The neural network is able to learn a large portion of the difference when evaluated offline on a validation set. When coupled to the low-resolution model, we find large forecast improvements up to one day on average. After this, the accumulated error due to the mass conservation violation of the neural network starts to dominate and deteriorates the forecast. This deterioration can effectively be delayed by adding a penalty term to the loss function used to train the ANN to conserve mass in a weak sense. This study reinforces the need to include physical constraints in neural network parameterizations.

Raphael Kriegmair et al.

Status: open (until 12 Jul 2021)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on npg-2021-20', Julien Brajard, 15 Jun 2021 reply

Raphael Kriegmair et al.

Raphael Kriegmair et al.

Viewed

Total article views: 254 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
222 29 3 254 0 1
  • HTML: 222
  • PDF: 29
  • XML: 3
  • Total: 254
  • BibTeX: 0
  • EndNote: 1
Views and downloads (calculated since 17 May 2021)
Cumulative views and downloads (calculated since 17 May 2021)

Viewed (geographical distribution)

Total article views: 235 (including HTML, PDF, and XML) Thereof 235 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 18 Jun 2021
Download
Short summary
Our regional numerical weather prediction models run at kilometer scale resolutions. Processes that occur at smaller scales are not resolved, yet significantly contribute to the atmospheric flow. We use a neural network (NN) to represent the unresolved part of physical process such as cumulus clouds. We test this approach on simplified, yet representative 1D model and find that the NN corrections vastly improve the model forecast up to a couple of days.