This page evaluates RRTMGP-NN, an accelerated version of the RRTMGP gas-optics module that replaces the computational kernel of the original scheme with neural networks, keeping the spectral resolution of RRTMGP. It predicts the optical depths in each spectral interval, and is therefore different from some other neural network radiation schemes that predict fluxes directly. It is designed for weather and climate applications. The model used in this evaluation takes a large number of minor gases as input (all 16 non-constant RRTMGP long-wave gases), 9 of these are not included in CKDMIP and were set to zero. The comparisons below use the 50 profiles of the "Evaluation-1" CKDMIP dataset. The reference calculations were performed using LBLRTM to generate the high resolution absorption spectra and the CKDMIP software to perform the radiative transfer.

Longwave

The longwave gas optics scheme uses a total of 256 g-points (k terms) in 16 bands. The evaluation has been performed using radiative transfer with four zenith angles in each hemisphere (8-stream).  The following plots evaluate fluxes and heating rates for the four CKDMIP greenhouse-gas scenarios "Present", "Preindustrial", "Glacial Maximum" and "Future". The results are very similar to those for the original RRTMGP scheme.

To examine the performance in more detail, the following plot evaluates fluxes and heating rates in each of the 13 CKDMIP longwave bands.

The following plot compares the instantaneous radiative forcing (change to net flux) at top-of-atmosphere and the surface, from perturbing the concentrations of individual well-mixed greenhouse gases from their present-day values, found by averaging over the 50 profiles of the Evaluation-1 dataset. For carbon dioxide (surface and TOA) and for methane (TOA), the results are excellent, similar to those for the original RRTMGP scheme. However, for the other gases the performance is not quite so good.

The following plot evaluates the representation of the overlap of the longwave absorption by carbon dioxide, methane and nitrous oxide. In each case, the x-axis shows the top-of-atmosphere radiative forcing from perturbing a gas to either its climatic minimum or maximum value, using the ranges stated by Hogan and Matricardi (2020). These radiative forcings are computed keeping the concentrations of all other well-mixed gases at their present-day values, except for the gas on the y-axis which is perturbed to its own climatic minimum or maximum values.