Physics-guided multi-task learning for subgrid scale turbulence parameterization: a comparative study of physics integration strategies

[1] S. B. Pope, Turbulent Flows. Cambridge, UK: Cambridge UniversityPress, 2000. [2] T. Schneider, J. Teixeira, C. S. Bretherton, F. Brient, K. G. Pressel, C. Schar, and A. P. Siebesma, “Climate goals and computing the future of clouds,” Nature Climate Change, vol. 7, no. 1, pp. 3–5, 2017. [3] A. R. Brown, R. T. Cederwall, A. Chlond, P. G. Duynkerke, J.-C. Golaz, M. Khairoutdinov, D. C. Lewellen, A. P. Lock, M. K. MacVean, C.-H. Moeng, et al., “Large-eddy simulation of the diurnal cycle of shallow cumulus convection over land,” Quarterly Journal of the Royal Meteorological Society, vol. 128, no. 582, pp. 1075–1093, 2002. [4] J. Smagorinsky, “General circulation experiments with the primitive equations: I. The basic experiment,” Monthly Weather Review, vol. 91, no. 3, pp. 99–164, 1963. [5] D. K. Lilly, “On the numerical simulation of buoyant convection,” Tellus, vol. 14, no. 2, pp. 148–172, 1962. [6] A. Pal, “Deep learning emulation of subgrid-scale processes in turbulent shear flows,” Geophysical Research Letters, vol. 47, no. 12, p. e2020GL087005, 2020. [7] D. Kochkov, J. A. Smith, A. Alieva, Q. Wang, M. P. Brenner, andS. Hoyer, “Machine learning–accelerated computational fluid dynamics,” Proceedings of the National Academy of Sciences, vol. 118, no. 21, p. e2101784118, 2021. [8] S. Rasp, M. S. Pritchard, and P. Gentine, “Deep learning to represent subgrid processes in climate models,” Proceedings of the National Academy of Sciences, vol. 115, no. 39, pp. 9684–9689, 2018. [9] H. Mandler and B. Weigand, “Generalization limits of data-driven turbulence models,” Flow, Turbulence and Combustion, 2024. [10] R. Molinaro, S. Lanthaler, B. Raonic, T. Rohner, V. Armegioiu, S. Simonis, D. Grund, Y. Ramic, Z. Y. Wan, F. Sha, S. Mishra, and L. ZepedaNu´nez, “Generative AI for fast and accurate statistical computation of fluids,” arXiv preprint arXiv:2409.18359, 2024. [11] M. Sakarvadia, K. Hegazy, A. Totounferoush, K. Chard, Y. Yang, I. Foster, and M. W. Mahoney, “The false promise of zeroshot super-resolution in machine-learned operators,” arXiv preprint arXiv:2510.06646, 2024. [12] G. E. Karniadakis, I. G. Kevrekidis, L. Lu, P. Perdikaris, S. Wang, and L. Yang, “Physics-informed machine learning,” Nature Reviews Physics, vol. 3, no. 6, pp. 422–440, 2021. [13] A. Kendall, Y. Gal, and R. Cipolla, “Multi-task learning using uncertainty to weigh losses for scene geometry and semantics,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 7482–7491, 2018. [14] R. Caruana, “Multitask learning,” Machine Learning, vol. 28, no. 1, pp. 41–75, 1997. [15] S. Liu, E. Johns, and A. J. Davison, “End-to-end multi-task learning with attention,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 1871–1880, 2019. [16] M. Raissi, P. Perdikaris, and G. E. Karniadakis, “Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations,” Journal of Computational Physics, vol. 378, pp. 686–707, 2019. [17] J. Yuval and P. A. O’Gorman, “Stable machine-learning parameterization of subgrid processes for climate modeling at a range of resolutions,” Nature Communications, vol. 11, no. 1, p. 3295, 2020. [18] T. Beucler, M. Pritchard, S. Rasp, J. Ott, P. Baldi, and P. Gentine, “Enforcing analytic constraints in neural networks emulating physical systems,” Physical Review Letters, vol. 126, no. 9, p. 098302, 2021. [19] H. V. Gupta, H. Kling, K. K. Yilmaz, and G. F. Martinez, “Decomposition of the mean squared error and NSE performance criteria: Implications for improving hydrological modelling,” Journal of Hydrology, vol. 377, no. 1-2, pp. 80–91, 2009.