Journal cover Journal topic
Geoscientific Model Development An interactive open-access journal of the European Geosciences Union
Journal topic

Journal metrics

Journal metrics

  • IF value: 5.154 IF 5.154
  • IF 5-year value: 5.697 IF 5-year
  • CiteScore value: 5.56 CiteScore
  • SNIP value: 1.761 SNIP 1.761
  • IPP value: 5.30 IPP 5.30
  • SJR value: 3.164 SJR 3.164
  • Scimago H <br class='hide-on-tablet hide-on-mobile'>index value: 59 Scimago H
    index 59
  • h5-index value: 49 h5-index 49
Discussion papers
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Submitted as: model description paper 21 Jan 2020

Submitted as: model description paper | 21 Jan 2020

Review status
This preprint is currently under review for the journal GMD.

RadNet 1.0: Exploring deep learning architectures for longwave radiative transfer

Ying Liu, Rodrigo Caballero, and Joy Merwin Monteiro Ying Liu et al.
  • Department of Meteorology, Stockholm University, Stockholm, Sweden

Abstract. Simulating global and regional climate at high resolution is essential to study the effects of climate change and capture extreme events affecting human populations. To achieve this goal, the scalability of climate models and the efficiency of individual model components are both important. Radiative transfer is among the most computationally expensive components in a typical climate model. Here we attempt to model this component using a neural network. We aim to study the feasibility of replacing an explicit, physics-based computation of longwave radiative transfer by a neural network emulator, and assessing the resultant performance gains. We compare multiple neural-network architectures, including a convolutional neural network and our results suggest that the performance loss from the use of convolutional networks is not offset by gains in accuracy. We train the networks with and without noise added to the input profiles and find that adding noise improves the ability of the networks to generalise beyond the training set. Prediction of radiative heating rates using our neural network models achieve up to 370x speedup on a GTX 1080 GPU setup and 11x speedup on a Xeon CPU setup compared to the a state of the art radiative transfer library running on the same Xeon CPU. Furthermore, our neural network models yield less than 0.1 Kelvin per day mean squared error across all pressure levels. Upon introducing this component into a single column model, we find that the time evolution of the temperature and humidity profiles are physically reasonable, though the model is conservative in its prediction of heating rates in regions where the optical depth changes quickly. Differences exist in the equilibrium climate simulated when using the neural networks, which are attributed to small systematic errors that accumulate over time. Thus, we find that the accuracy of the neural network in the "offline" mode does not reflect its performance when coupled with other components.

Ying Liu et al.

Interactive discussion

Status: open (until 17 Mar 2020)
Status: open (until 17 Mar 2020)
AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment
[Subscribe to comment alert] Printer-friendly Version - Printer-friendly version Supplement - Supplement

Ying Liu et al.

Ying Liu et al.


Total article views: 190 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
149 34 7 190 6 5
  • HTML: 149
  • PDF: 34
  • XML: 7
  • Total: 190
  • BibTeX: 6
  • EndNote: 5
Views and downloads (calculated since 21 Jan 2020)
Cumulative views and downloads (calculated since 21 Jan 2020)

Viewed (geographical distribution)

Total article views: 139 (including HTML, PDF, and XML) Thereof 139 with geography defined and 0 with unknown origin.
Country # Views %
  • 1



No saved metrics found.


No discussed metrics found.
Latest update: 28 Feb 2020
Publications Copernicus
Short summary
The calculation of atmospheric radiative transfer is the most computationally expensive part of climate models. Reducing this computational burden could potentially improve our ability to simulate the earth’s climate at finer scales. We propose using a statistical model — a deep neural network — to compute approximate radiative transfer in the earth’s atmosphere. We demonstrate a significant reduction in computational burden as compared to a traditional scheme, especially when using GPUs.
The calculation of atmospheric radiative transfer is the most computationally expensive part of...