Preprints
https://doi.org/10.5194/egusphere-2022-1047
https://doi.org/10.5194/egusphere-2022-1047
 
09 Dec 2022
09 Dec 2022
Status: this preprint is open for discussion.

Implementation of a machine-learned gas optics parameterization in the ECMWF Integrated Forecasting System: RRTMGP-NN 2.0

Peter Ukkonen1 and Robin Hogan2,3 Peter Ukkonen and Robin Hogan
  • 1Danish Meteorological Institute
  • 2European Centre for Medium-Range Weather Forecasts, Reading, UK
  • 3Department of Meteorology, University of Reading, Reading, UK

Abstract. Radiation schemes are physically important but computationally expensive components of weather and climate models. This has spurred efforts to replace them with a cheap emulator based on neural networks (NN), obtaining large speed-ups, but at the expense of accuracy, energy conservation and generalization. An alternative approach which is slower but more robust than full emulation is to use NNs to predict optical properties, but keep the radiative transfer equations. Recently, NNs were developed to replace the RRTMGP gas optics scheme, and shown to be accurate while improving speed.However, the evaluations were based solely on offline radiation computations.

In this paper, we describe the implementation and prognostic evaluation of RRTMGP-NN in the Integrated Forecasting System (IFS) of the European Centre for Medium-Range Weather Forecasts (ECMWF). The new gas optics scheme was incorporated into ecRad, the modular ECMWF radiation scheme. Using a hybrid loss function designed to reduce radiative forcing errors, and an early stopping method based on monitoring fluxes and heating rates with respect to a line-by-line benchmark, new NN models were trained on RRTMGP k-distributions with reduced spectral resolutions. Offline evaluation of the new NN gas optics, RRTMGP-NN 2.0, shows a very high level of accuracy for clear-sky fluxes and heating rates; for instance the RMSE in shortwave surface downwelling flux is 0.78 W m−2 for RRTMGP and 0.80 W m−2 for RRTMGP-NN in a present-day scenario, while upwelling flux errors are actually smaller for the NN. Because our approach does not affect the treatment of clouds, no additional errors will be introduced for cloudy profiles. RRTMGP-NN closely reproduces radiative forcings for 5 important greenhouse gases across a wide range of concentrations such as 8x CO2.

To assess the impact of different gas optics schemes in the IFS, four 1-year coupled ocean-atmosphere simulations were performed for each configuration. The results show that RRTMGP-NN and RRTMGP produce very similar model climates, with the differences being smaller than those between existing schemes, and statistically insignificant for zonal means of single-level quantities such as surface temperature. The use of RRTMGP-NN speeds up ecRad by a factor of 1.5 compared to RRTMGP (the gas optics being almost 3 times faster), and is also faster than the older and less accurate RRTMG which is used in the current operational cycle of the IFS.

Peter Ukkonen and Robin Hogan

Status: open (until 18 Feb 2023)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse

Peter Ukkonen and Robin Hogan

Peter Ukkonen and Robin Hogan

Viewed

Total article views: 200 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
146 46 8 200 3 3
  • HTML: 146
  • PDF: 46
  • XML: 8
  • Total: 200
  • BibTeX: 3
  • EndNote: 3
Views and downloads (calculated since 09 Dec 2022)
Cumulative views and downloads (calculated since 09 Dec 2022)

Viewed (geographical distribution)

Total article views: 227 (including HTML, PDF, and XML) Thereof 227 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 03 Feb 2023
Download
Short summary
Climate and weather models suffer from uncertainties resulting from approximated processes. Solar and thermal radiation is one example, as it’s computationally too costly to simulate precisely. This has led to attempts to replace radiation codes using physical equations with neural networks (NNs), that are faster but highly uncertain. In this paper we use global weather simulations to demonstrate that a middle-ground approach of using NNs to predict optical properties is much more accurate.