Abstract
Dust sharply degrades visibility in arid and semi-arid regions, yet operational forecasting remains challenged by near-surface process errors in numerical weather prediction (NWP) and the poor generalization of purely data-driven models. We present a physics-guided machine learning (PGML) framework that post-processes European Centre for Medium-Range Weather Forecasts (ECMWF) forecasts to predict five ordinal visibility grades over the Kumtag Desert. A dust-lifecycle feature library (emission, vertical mixing, transport, wet scavenging) is coupled with an ordinal LightGBM architecture. On an independent test period, the model attains quadratic weighted kappa (QWK) of 0.26 (0–24 h), 0.17 (24–48 h), and 0.18 (48–72 h), with mean absolute error (MAE) of 0.48–0.56; gains versus data-only baselines increase with forecast horizon. Ablation experiments show that physics priors can effectively improve visibility prediction accuracy—reducing MAE by up to 10% and sustaining QWK beyond 24 h by constraining non-physical drift. Accordingly, the PGML visibility predictions show improved performance relative to data-only baselines. SHAP analysis reveals a forecast-horizon-dependent mechanistic shift from emission/surface-layer dynamics to stability-controlled vertical mixing, consistent with dust dynamics. The framework offers an interpretable, transferable paradigm for physics-constrained environmental forecasting.
Similar content being viewed by others
Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
References
Veal, A. J. Climate change 2021: the physical science basis, 6th report. World Leisure J. 63, 443–444 (2021). https://doi.org/10.1080/16078055.2021.2008646
Kok, J. F. et al. Mineral dust aerosol impacts on global climate and climate change. Nat. Rev. Earth Environ. 4, 71–86. https://doi.org/10.1038/s43017-022-00379-5 (2023).
Prospero, J. M., Ginoux, P., Torres, O., Nicholson, S. E. & Gill, T. E. Environmental characterization of global sources of atmospheric soil dust identified with the nimbus 7 total Ozone mapping spectrometer (TOMS) absorbing aerosol product. Rev. Geophys. 40, 31. https://doi.org/10.1029/2000rg000095 (2002).
Wang, G. C., Shu, S. J. & Li, W. J. Asian dust threatens air pollution control efforts. Science 390, 3. https://doi.org/10.1126/science.aeb2629 (2025).
Malm, W. C. & Hand, J. L. An examination of the physical and optical properties of aerosols collected in the IMPROVE program. Atmos. Environ. 41, 3407–3427. https://doi.org/10.1016/j.atmosenv.2006.12.012 (2007).
Dubovik, O. et al. Variability of absorption and optical properties of key aerosol types observed in worldwide locations. J. Atmos. Sci. 59, 590–608. https://doi.org/10.1175/1520-0469(2002)059%3C0590:Voaaop%3E2.0.Co;2 (2002).
Benedetti, A. et al. Status and future of numerical atmospheric aerosol prediction with a focus on data requirements. Atmos. Chem. Phys. 18, 10615–10643. https://doi.org/10.5194/acp-18-10615-2018 (2018).
Gong, S. L. et al. Characterization of soil dust aerosol in China and its transport and distribution during 2001 ACE-Asia: 2. Model simulation and validation - art. 4262. J. Geophys. Res. -Atmos. 108, 19. https://doi.org/10.1029/2002jd002633 (2003).
Luan, T., Guo, X. L., Guo, L. J. & Zhang, T. H. Quantifying the relationship between PM < sub>2.5 concentration, visibility and planetary boundary layer height for long-lasting haze and fog-haze mixed events in Beijing. Atmos. Chem. Phys. 18, 203–225. https://doi.org/10.5194/acp-18-203-2018 (2018).
Song, J. I., Yum, S. S., Gultepe, I., Chang, K. H. & Kim, B. G. Development of a new visibility parameterization based on the measurement of fog microphysics at a mountain site in Korea. Atmos. Res. 229, 115–126. https://doi.org/10.1016/j.atmosres.2019.06.011 (2019).
Hu, S. Y. et al. Current challenges of improving visibility due to increasing nitrate fraction in PM < sub>2.5 during the haze days in Beijing, China. Environ. Pollut. 290, 8. https://doi.org/10.1016/j.envpol.2021.118032 (2021).
Zhou, C. H. et al. Detection of new dust sources in Central/East Asia and their impact on simulations of a severe sand and dust storm. J. Geophys. Res. -Atmos. 124, 10232–10247. https://doi.org/10.1029/2019jd030753 (2019).
Karagulian, F. et al. Analysis of a severe dust storm and its impact on air quality conditions using WRF-Chem modeling, satellite imagery, and ground observations. Air Qual. Atmos. Health. 12, 453–470. https://doi.org/10.1007/s11869-019-00674-z (2019).
Wang, Y. Q. et al. Surface observation of sand and dust storm in East Asia and its application in CUACE/Dust. Atmos. Chem. Phys. 8, 545–553. https://doi.org/10.5194/acp-8-545-2008 (2008).
Bi, K. F. et al. Accurate medium-range global weather forecasting with 3D neural networks. Nature 619, 533–. https://doi.org/10.1038/s41586-023-06185-3 (2023).
Price, I. et al. Probabilistic weather forecasting with machine learning. Nature 637, 21. https://doi.org/10.1038/s41586-024-08252-9 (2025).
Chen, L. et al. FuXi: a cascade machine learning forecasting system for 15-day global weather forecast. Npj Clim. Atmos. Sci. 6, 11. https://doi.org/10.1038/s41612-023-00512-1 (2023).
Ortega, L. C., Otero, L. D., Solomon, M., Otero, C. E. & Fabregas, A. Deep learning models for visibility forecasting using Climatological data. Int. J. Forecast. 39, 992–1004. https://doi.org/10.1016/j.ijforecast.2022.03.009 (2023).
Peláez-Rodríguez, C. et al. Deep learning ensembles for accurate fog-related low-visibility events forecasting. Neurocomputing 549 https://doi.org/10.1016/j.neucom.2023.126435 (2023).
Penov, N. & Guerova, G. Sofia airport visibility Estimation with two Machine-Learning techniques. Remote Sens. 15 https://doi.org/10.3390/rs15194799 (2023).
Shankar, A. & Sahana, B. C. Early warning of low visibility using the ensembling of machine learning approaches for aviation services at Jay Prakash Narayan international (JPNI) airport Patna. SN Appl. Sci. 5 https://doi.org/10.1007/s42452-023-05350-7 (2023).
Zhang, Y. et al. Visibility prediction based on machine learning algorithms. Atmosphere 13 https://doi.org/10.3390/atmos13071125 (2022).
Karniadakis, G. E. et al. Physics-informed machine learning. Nat. Rev. Phys. 3, 422–440. https://doi.org/10.1038/s42254-021-00314-5 (2021).
Kashinath, K. et al. Physics-informed machine learning: case studies for weather and climate modelling. Philos. Trans. R Soc. A-Math Phys. Eng. Sci. 379, 36. https://doi.org/10.1098/rsta.2020.0093 (2021).
Liu, W., Lai, Z. L., Bacsa, K. & Chatzi, E. Physics-guided deep Markov models for learning nonlinear dynamical systems with uncertainty. Mech. Syst. Signal. Proc. 178 (20). https://doi.org/10.1016/j.ymssp.2022.109276 (2022).
Jin, J. B. et al. Source Backtracking for dust storm emission inversion using an adjoint method: case study of Northeast China. Atmos. Chem. Phys. 20, 15207–15225. https://doi.org/10.5194/acp-20-15207-2020 (2020).
Kok, J. F., Albani, S., Mahowald, N. M. & Ward, D. S. An improved dust emission model - Part 2: evaluation in the community Earth system Model, with implications for the use of dust source functions. Atmos. Chem. Phys. 14, 13043–13061. https://doi.org/10.5194/acp-14-13043-2014 (2014).
Morcrette, J. J. et al. Aerosol analysis and forecast in the European centre for Medium-Range weather forecasts integrated forecast system: forward modeling. J. Geophys. Res. -Atmos. 114, 17. https://doi.org/10.1029/2008jd011235 (2009).
Ding, C., Feng, G. C., Zhang, L. & Liao, M. S. A novel multidimensional perspective on dynamic characteristics of the peculiar Feather-Shaped dunes in Kumtag desert with Time-Series optical and SAR observations. IEEE J. Sel. Top. Appl. Earth Observ Remote Sens. 17, 11618–11631. https://doi.org/10.1109/jstars.2024.3414449 (2024).
Eck, T. F. et al. Columnar aerosol optical properties at AERONET sites in central Eastern Asia and aerosol transport to the tropical mid-Pacific - art. no. D06202. J. Geophys. Res. -Atmos. 110, 18. https://doi.org/10.1029/2004jd005274 (2005).
Ginoux, P. et al. Rev. Geophys. 50, 36 https://doi.org/10.1029/2012rg000388 (2012).
Swana, E. F., Doorsamy, W. & Bokoro, P. Tomek link and SMOTE approaches for machine fault classification with an imbalanced dataset. Sensors 22, 21. https://doi.org/10.3390/s22093246 (2022).
Li, G., Zhang, J., Herrmann, H. J., Shao, Y. & Huang, N. Study of aerodynamic grain entrainment in aeolian transport. Geophys. Res. Lett. 47 https://doi.org/10.1029/2019GL086574 (2020).
Marticorena, B. & Bergametti, G. Modeling the atmospheric dust cycle: 1. Design of a soil-derived dust emission scheme. J. Geophys. Res. Atmos. 100, 16415–16430. https://doi.org/10.1029/95JD00690 (1995).
Anderson, R. S. & Haff, P. K. Wind modification and bed response during saltation of sand in air. Acta Mech. Suppl. (1), 21–51. https://doi.org/10.1007/978-3-7091-6706-9_2 (1991).
Shao, Y. & Lu, H. A simple expression for wind erosion threshold friction velocity. J. Geophys. Res. Atmos. 105, 22437–22443. https://doi.org/10.1029/2000JD900304 (2000).
Zhang, H. W. et al. Numerical simulation of wind field and sand flux in crescentic sand dunes. Sci. Rep. 11, 18. https://doi.org/10.1038/s41598-021-84509-x (2021).
Khalfallah, B. et al. Influence of atmospheric stability on the size distribution of the vertical dust flux measured in eroding conditions over a flat bare sandy field. J. Geophys. Res. -Atmos. 125, 20. https://doi.org/10.1029/2019jd031185 (2020).
Knippertz, P., Todd, M. C., MINERAL DUST AEROSOLS OVER THE & SAHARA: METEOROLOGICAL CONTROLS ON EMISSION AND TRANSPORT AND IMPLICATIONS FOR MODELING. Rev. Geophys. 50, 28 https://doi.org/10.1029/2011rg000362 (2012).
Giannakopoulou, E. M. & Toumi, R. The Persian Gulf summertime low-level jet over sloping terrain. Q. J. R Meteorol. Soc. 138, 145–157. https://doi.org/10.1002/qj.901 (2012).
Heinold, B., Tegen, I., Schepanski, K. & Hellmuth, O. Dust radiative feedback on saharan boundary layer dynamics and dust mobilization. Geophys. Res. Lett. 35, 5. https://doi.org/10.1029/2008gl035319 (2008).
Yu, Z., Ma, J., Qu, Y., Pan, L. & Wan, S. PM2.5 extended-range forecast based on MJO and S2S using LightGBM. Sci. Total Environ. 880 https://doi.org/10.1016/j.scitotenv.2023.163358 (2023).
Huang, J. et al. Taklimakan dust aerosol radiative heating derived from CALIPSO observations using the Fu-Liou radiation model with CERES constraints. Atmos. Chem. Phys. 9, 4011–4021. https://doi.org/10.5194/acp-9-4011-2009 (2009).
Su, T. N. et al. An intercomparison of long-term planetary boundary layer heights retrieved from CALIPSO, ground-based lidar, and radiosonde measurements over Hong Kong. J. Geophys. Res. -Atmos. 122, 3929–3943. https://doi.org/10.1002/2016jd025937 (2017).
Li, Y. R. et al. Long-term variation of boundary layer height and possible contribution factors: A global analysis. Sci. Total Environ. 796, 14. https://doi.org/10.1016/j.scitotenv.2021.148950 (2021).
Xu, C., Ma, Y. M., Yang, K. & You, C. Tibetan plateau impacts on global dust transport in the upper troposphere. J. Clim. 31, 4745–4756. https://doi.org/10.1175/jcli-d-17-0313.1 (2018).
Adams, A. M., Prospero, J. M. & Zhang, C. D. CALIPSO-Derived Three-Dimensional structure of aerosol over the Atlantic basin and adjacent continents. J. Clim. 25, 6862–6879. https://doi.org/10.1175/jcli-d-11-00672.1 (2012).
Chouza, F., Reitebuch, O., Benedetti, A. & Weinzierl, B. Saharan dust long-range transport across the Atlantic studied by an airborne doppler wind lidar and the MACC model. Atmos. Chem. Phys. 16, 11581–11600. https://doi.org/10.5194/acp-16-11581-2016 (2016).
Zhou, B., Liu, D. Y. & Yan, W. L. A simple new method for calculating precipitation scavenging effect on particulate matter: based on Five-Year data in Eastern China. Atmosphere 12, 12. https://doi.org/10.3390/atmos12060759 (2021).
Abdelkader, M. et al. Dust-air pollution dynamics over the Eastern mediterranean. Atmos. Chem. Phys. 15, 9173–9189. https://doi.org/10.5194/acp-15-9173-2015 (2015).
Morales-Martín, A. et al. Deep ordinal classification in forest areas using light detection and ranging point clouds. Sensors 24, 18. https://doi.org/10.3390/s24072168 (2024).
Acknowledgements
We would like to express our sincere gratitude to the Editor and the anonymous reviewers for their careful evaluation and constructive comments, which have significantly improved the quality, clarity, and rigor of this manuscript.
Funding
This research was funded by the Graduate Education Innovation Plan Project of the Education Department of Xinjiang Uygur Autonomous Region (Grant No. XJ2024G083), and supported by the Taishan Industrial Experts Program.
Author information
Authors and Affiliations
Contributions
C.X. and H.Z. contributed to methodology and investigation. J.L., Y.S., and H.Q. provided resources. C.X. and H.Z. wrote the original draft. C.Z., K.L. Y.Q., and H.Z. reviewed and edited the manuscript. C.Z., H.Z., and C.X. contributed to visualization. Z. H, C.X. and H.Z. made contributions in revising the manuscript. All authors read and approved the final manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix A: parameter settings
Appendix A: parameter settings
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Xu, C., Zhang, H., Luo, K. et al. A physics-guided machine learning framework for enhancing dust storm visibility prediction in arid and semi-arid regions. Sci Rep (2026). https://doi.org/10.1038/s41598-026-39766-z
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-026-39766-z
