&amp;lt;TT&amp;gt;sympl&amp;lt;/TT&amp;gt; (v. 0.3.2) and &amp;lt;TT&amp;gt;climt&amp;lt;/TT&amp;gt; (v. 0.11.0) – Towards a flexible
framework for building model hierarchies in Python

Monteiro, Joy Merwin; McGibbon, Jeremy; Caballero, Rodrigo

doi:10.5194/gmd-2018-100

Cited by 6 publications

(2 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The calculation of the radiative fields for the historical and perturbed datasets are calculated using the RRTMG component available in the climt modelling toolkit (Monteiro and Caballero, 2016;Monteiro et al, 2018). This component is a python wrapper over the RRTMG fortran library, and provides convenient access to the radiation fields.…”

Section: The Rt Datasetmentioning

confidence: 99%

“…To limit the scope of this exploratory study, we focus on longwave radiative transfer under clear-sky conditions (henceforth, RT thus refers to clear-sky longwave radiative transfer). We use the RRTMG library available within the climt modelling toolkit (Monteiro et al, 2018) to generate radiative cooling profiles to train the NN models. In particular, we compare the accuracy-computational complexity tradeoff between five kinds of NN architectures on both CPU and GPU.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

RadNet 1.0: Exploring deep learning architectures for longwave radiative transfer

Liu¹,

Caballero²,

Monteiro³

2020

Preprint

Self Cite

View full text Add to dashboard Cite

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.

show abstract

Section: The Rt Datasetmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%