A Physics‐Incorporated Deep Learning Framework for Parameterization of Atmospheric Radiative Transfer.

The atmospheric radiative transfer calculations are among the most time‐consuming components of the numerical weather prediction (NWP) models. Deep learning (DL) models have recently been increasingly applied to accelerate radiative transfer modeling. Besides, a physical relationship exists between the output variables, including fluxes and heating rate profiles. Integration of such physical laws in DL models is crucial for the consistency and credibility of the DL‐based parameterizations. Therefore, we propose a physics‐incorporated framework for the radiative transfer DL model, in which the physical relationship between fluxes and heating rates is encoded as a layer of the network so that the energy conservation can be satisfied. It is also found that the prediction accuracy was improved with the physic‐incorporated layer. In addition, we trained and compared various types of DL model architectures, including fully connected (FC) neural networks (NNs), convolutional‐based NNs (CNNs), bidirectional recurrent‐based NNs (RNNs), transformer‐based NNs, and neural operator networks, respectively. The offline evaluation demonstrates that bidirectional RNNs, transformer‐based NNs, and neural operator networks significantly outperform the FC NNs and CNNs due to their capability of global perception. A global perspective of an entire atmospheric column is essential and suitable for radiative transfer modeling as the changes in atmospheric components of one layer/level have both local and global impacts on radiation along the entire vertical column. Furthermore, the bidirectional RNNs achieve the best performance as they can extract information from both upward and downward directions, similar to the radiative transfer processes in the atmosphere. Plain Language Summary: Numerical weather prediction models require a lot of computational resources and time to run. Calculating the atmospheric radiative transfer processes is one of the most computationally expensive parts of the NWP model. One alternative is to model the radiative transfer using deep learning (DL) models, but the DL models do not involve physical laws and may have physically inconsistent outputs. This paper proposes a DL model framework to ensure the thermal equilibrium between fluxes and heating rates, which are outputs of radiative transfer models. Also, the accuracy of DL‐based radiative transfer prediction is improved when using the framework. Various DL models have been trained and compared. The results demonstrate that model structures with global receptive fields work best for emulating radiative transfer calculations. Key Points: A physics‐incorporated deep learning (DL) framework for parameterization of atmospheric radiative transfer is proposedThe DL model structures with global receptive fields are more suitable for the radiative transfer problem [ABSTRACT FROM AUTHOR]