Abstract
Correcting the forecast bias of numerical weather prediction models is important for severe weather warnings. The refined grid forecast requires direct correction on gridded forecast products, as opposed to correcting forecast data only at individual weather stations. In this study, a deep learning method called CU-net is proposed to correct the gridded forecasts of four weather variables from the European Centre for Medium-Range Weather Forecast Integrated Forecasting System global model (ECMWF-IFS): 2-m temperature, 2-m relative humidity, 10-m wind speed, and 10-m wind direction, with a forecast lead time of 24 h to 240 h in North China. First, the forecast correction problem is transformed into an image-to-image translation problem in deep learning under the CU-net architecture, which is based on convolutional neural networks. Second, the ECMWF-IFS forecasts and ECMWF reanalysis data (ERA5) from 2005 to 2018 are used as training, validation, and testing datasets. The predictors and labels (ground truth) of the model are created using the ECMWF-IFS and ERA5, respectively. Finally, the correction performance of CU-net is compared with a conventional method, anomaly numerical correction with observations (ANO). Results show that forecasts from CU-net have lower root mean square error, bias, mean absolute error, and higher correlation coefficient than those from ANO for all forecast lead times from 24 h to 240 h. CU-net improves upon the ECMWF-IFS forecast for all four weather variables in terms of the above evaluation metrics, whereas ANO improves upon ECMWF-IFS performance only for 2-m temperature and relative humidity. For the correction of the 10-m wind direction forecast, which is often difficult to achieve, CU-net also improves the correction performance.
Trích muốn
Đối số giá trị dự báo thời tiết (NWP) tiến hành lệch lạc đính chính, là đề cao NWP kết quả chuẩn xác tính cùng nghiệp vụ chống đỡ năng lực quan trọng con đường. Cùng truyền thống trạm điểm dự báo kết quả đính chính so sánh với, trước mắt tinh tế hóa dự báo yêu cầu đối NWP cách điểm kết quả trực tiếp tiến hành đính chính. Bổn nghiên cứu đưa ra một loại chiều sâu học tập phương pháp CU-net, cùng sử dụng này đối Châu Âu trung kỳ dự báo thời tiết trung tâm toàn cầu trị số dự báo hệ thống (ECMWF-IFS) 2 mễ độ ấm, 2 mễ độ ẩm tương đối, 10 mễ phong 24-240 giờ dự báo tiến hành cách điểm đính chính. Đầu tiên, đem NWP lệch lạc đính chính vấn đề thay đổi vì “Hình ảnh đến hình ảnh” khí tượng tin tức phiên dịch vấn đề, cũng thông qua xây dựng chiều sâu học tập mô hình CU-net tiến hành cầu giải. Sau đó, sử dụng 2005-2018 năm ECMWF-IFS dự báo cùng ECMWF đệ 5 đại lại phân tích tư liệu (ECMWF-ERA5) cấu thành số liệu tập đối CU-net mô hình tiến hành huấn luyện, nghiệm chứng cùng thí nghiệm. Thí nghiệm kết quả cho thấy, cùng truyền thống hình thức cự bình tích phân dự báo đính chính phương pháp (ANO) so sánh với, CU-net ở đều căn thức khác biệt, lệch lạc, bình quân tuyệt đối khác biệt cùng tương quan hệ số chờ kiểm nghiệm chỉ tiêu thượng, đều lấy được càng ưu kết quả. Đối khí tượng lĩnh vực công nhận so khó đính chính 10 mễ tốc độ gió cùng hướng gió hai cái lượng biến đổi, CU-net đính chính tính năng cũng thực lộ rõ. Bổn văn nghiên cứu chỉ ra, chiều sâu học tập phương pháp có từ NWP rộng lượng số liệu trung trực tiếp tiến hành “Học tập” do đó xây dựng NWP lệch lạc đặc thù năng lực, vì NWP lệch lạc đính chính nghiên cứu cùng nghiệp vụ ứng dụng sáng lập tân con đường.
Article PDF
Similar content being viewed by others
Explore related subjects
Discover the latest articles, news and stories from top researchers in related subjects.Avoid common mistakes on your manuscript.
References
Bao, L., T. Gneiting, E. P. Grimit, P. Guttorp, and A. E. Raftery, 2010: Bias correction and Bayesian model averaging for ensemble forecasts of surface wind direction.Mon. Wea. Rev.,138,1811–1821,https://doi.org/10.1175/2009MWR3138.1.
Bauer, P., A. Thorpe, and G. Brunet, 2015: The quiet revolution of numerical weather prediction.Nature,525,47–55,https://doi.org/10.1038/nature14956.
Boukabara, S. A., V. Krasnopolsky, J. Q. Stewart, E. S. Maddy, N. Shahroudi, and R. N. Hoffman, 2019: Leveraging modern artificial intelligence for remote sensing and NWP: Benefits and challenges.Bull. Amer. Meteor. Soc.,100,ES473–ES491,https://doi.org/10.1175/BAMS-D-18-0324.1.
Chang, J., X. D. Peng, G. Z. Fan, and Y. Z. Che, 2015: Error correction of numerical weather prediction with historical data.Acta Meteorologica Sinica,73,341–354,https://doi.org/10.11676/qxxb2015.021.(in Chinese with English abstract)
Chen, H. N., V. Chandrasekar, H. M. Tan, and R. Cifelli, 2019: Rainfall estimation from ground radar and TRMM Precipitation Radar using hybrid deep neural networks.Geophys. Res. Lett.,46,10669–10678,https://doi.org/10.1029/2019GL084771.
Cho, D., C. Yoo, J. Im, and D. H. Cha, 2020: Comparative assessment of various machine learning-based bias correction methods for numerical weather prediction model forecasts of extreme air temperatures in urban areas.Earth and Space Science,7,e2019EA000740,https://doi.org/10.1029/2019EA000740.
Glahn, H. R., and D. A. Lowry, 1972: The use of model output statistics (MOS) in objective weather forecasting.J. Appl. Meteorol. Climatol.,11,1203–1211,https://doi.org/10.1175/1520-0450(1972)011<1203:TUOMOS>2.0.CO;2.
Guo, H. Y., M. X. Chen, and L. Han. 2020: Evaluation of the Conv-GRU deep learning method for convective weather nowcasting.Proc. 19th Conf. on Artificial Intelligence for Environmental Science,Boston, MA, Amer. Meteor. Soc.
Han, L., J. Z. Sun, and W. Zhang, 2020: Convolutional neural network for convective storm nowcasting using 3-D Doppler weather radar data.IEEE Trans. Geosci. Remote Sens.,58,1487–1495,https://doi.org/10.1109/TGRS.2019.2948070.
He, D. X., Z. M. Zhou, Z. P. Kang, and L. Liu, 2019: Numerical studies on forecast error correction of GRAPES model with variational approach.Advances in Meteorology,2019,2856289,https://doi.org/10.1155/2019/2856289.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis.Quart. J. Roy. Meteor. Soc.,146,1999–2049,https://doi.org/10.1002/qj.3803.
Homleid, M., 1995: Diurnal corrections of short-term surface temperature Forecasts Using the Kalman Filter.Wea. Forecasting,10,689–707,https://doi.org/10.1177/1520-0434(1995)010<0689:DCOSTS>2.0.CO;2.
Houtekamer, P. L., and H. L. Mitchell, 1998: Data assimilation using an ensemble Kalman filter technique.Mon. Wea. Rev.,126,796–811,https://doi.org/10.1175/1520-0493(1998)126<0796:DAUAEK>2.0.CO;2.
Hu, S. J., C. Y. Qiu, L. Y. Zhang, Q. C. Huang, H. P. Yu, and J. F. Chou, 2014: An approach to estimating and extrapolating model error based on inverse problem methods: Towards accurate numerical weather prediction.Chinese Physics B,23,089201,https://doi.org/10.1088/1674-1056/23/8/089201.
Kingma, D. P., and J. Ba, 2015: Adam: A Method for Stochastic Optimization.Proc. 3rd International Conf. on Learning Representations,San Diego, ICLR.
Klein, W. H., B. M. Lewis, and I. Enger, 1959: Objective prediction of five-day mean temperatures during winter.J. Atmos. Sci.,16,672–682,https://doi.org/10.1175/1520-0469(1959)016<0672:OPOFDM>2.0.CO;2.
Lagerquist, R., A. McGovern, and D. J. Gagne II, 2019: Deep learning for spatially explicit prediction of synoptic-scale Fronts.Wea. Forecasting,34,1137–1160,https://doi.org/10.1175/WAF-D-18-0183.1.
Lebedev, V., and Coauthors, 2019: Precipitation nowcasting with satellite imagery.Proc. 25th ACM SIGKDD Int. Conf. on Knowledge Discovery & Data Mining,Anchorage, AK, ACM, 2680–2688,https://doi.org/10.1145/3292500.3330762.
Marzban, C., 2003: Neural networks for postprocessing model output: ARPS.Mon. Wea. Rev.,131,1103–1111,https://doi.org/10.1175/1520-0493(2003)131<1103:NNFPMO>2.0.CO;2.
Marzban, C., S. Sandgathe, and E. Kalnay, 2006: MOS, perfect prog, and reanalysis.Mon. Wea. Rev.,134,657–663,https://doi.org/10.1175/MWR3088.1.
Nair, V., and Hinton, G. E., 2010: Rectified linear units improve restricted Boltzmann machines.Proc. 27th Int. Conf. on Machine Learning,Haifa, Israel, ICML, 807.
Peng, X. D., Y. Z. Che, and J. Chang, 2013: A novel approach to improve numerical weather prediction skills by using anomaly integration and historical data.J. Geophys. Res.,118,8814–8826,https://doi.org/10.1002/jgrd.50682.
Qian, W. H., 2012: How to improve the skills of weather and climate predictions?Chinese Journal of Geophysics,55,1532–1540,https://doi.org/10.6038/j.issn.0001-5733.2012.05.010.(in Chinese with English abstract)
Rasp, S., and S. Lerch, 2018: Neural networks for postprocessing ensemble weather forecasts.Mon. Wea. Rev.,146,3885–3900,https://doi.org/10.1175/MWR-D-18-0187.1.
Ronneberger, O., P. Fischer, and T. Brox, 2015: U-net: Convolutional networks for biomedical image segmentation.Proc. 18th Int. Conf. on Medical Image Computing and Computer-Assisted Intervention,Munich, Germany, Springer,https://doi.org/10.1007/978-3-319-24574-4_28.
Shi, W. Z., J. Caballero, F. Huszár, J. Totz, A. P. Aitken, R. Bishop, D. Rueckert, and Z. H. Wang, 2016: Real-time single image and video super-resolution using an efficient sub-pixel convolutional neural network.Proc. 2016 IEEE Conf. on Computer Vision and Pattern Recognition,Las Vegas, NV, IEEE,https://doi.org/10.1109/CVPR.2016.207.
Shi, X. J., Z. R. Chen, H. Wang, D. Y. Yeung, W. K. Wong, and W. C. Woo, 2015: Convolutional LSTM network: A machine learning approach for precipitation nowcasting.Proc. 28th Int. Conf. on Neural Information Processing Systems,Montreal, Quebec, NIPS, 802–810,https://doi.org/10.5555/2969239.2969329.
Shi, X. J., Z. H. Gao, L. Lausen, H. Wang, D. Y. Yeung, W. K. Wong, and W. C. Woo, 2017: Deep learning for precipitation nowcasting: A benchmark and a new model.Proc. 31st Int. Conf. on Neural Information Processing Systems,Long Beach, CA, NIPS, 5622–5632,https://doi.org/10.5555/3295222.3295313.
Steinacker, R., and Coauthors, 2006: A mesoscale data analysis and downscaling method over complex terrain.Mon. Wea. Rev.,134,2758–2771,https://doi.org/10.1175/MWR3196.1.
Tao, Y. M., X. G. Gao, K. Hsu, S. Sorooshian, and A. Ihler, 2016: A deep neural network modeling framework to reduce bias in satellite precipitation products.Journal of Hydrometeorology,17,931–945,https://doi.org/10.1175/JHM-D-15-0075.1.
Vannitsem, S., and Coauthors, 2020: Statistical postprocessing for weather forecasts—Review, challenges and avenues in a big data world. arXiv: 2004.06582v1.
Vashani, S., M. Azadi, and S. Hajjam, 2010: Comparative evaluation of different post processing methods for numerical prediction of temperature forecasts over Iran.Research Journal of Environmental Sciences,4,305–316,https://doi.org/10.3923/rjes.2010.305.316.
Vislocky, R. L., and G. S. Young, 1989: The use of perfect prog forecasts to improve model output statistics forecasts of precipitation probability.Wea. Forecasting,4,202–209,https://doi.org/10.1175/1520-0434(1989)004<0202:TUOPPF>2.0.CO;2.
Xia, J. J., and Coauthors, 2020: Machine learning-based weather support for the 2022 Winter Olympics.Adv. Atmos. Sci.,37,927–932,https://doi.org/10.1007/s00376-020-0043-5.
Xue, H. L., X. S. Shen, and J. F. Chou, 2015: An online model correction method based on an inverse problem: Part I—model error estimation by Iteration.Adv. Atmos. Sci.,2,1329–1340,https://doi.org/10.1007/s00376-015-4261-1.
Zeiler, M. D., and R. Fergus, 2014: Visualizing and understanding convolutional networks.Proc. 13th European Conf. on Computer Vision,Zurich, Switzerland, Springer, 818–833,https://doi.org/10.1007/978-3-319-10590-1_53.
Acknowledgments
This work was supported in part by the National Key R&D Program of China (Grant No. 2018YFF0300102), the National Natural Science Foundation of China (Grant Nos. 41875049 and 41575050), and the Beijing Natural Science Foundation (Grant No. 8212025). We gratefully acknowledge the support of NVIDIA Corporation for the donation of the GPU used for this research.
Author information
Authors and Affiliations
Corresponding author
Additional information
Article Highlights
• A deep learning method (CU-net) is proposed to correct gridded forecast products.
• CU-net demonstrates superior performance in correcting ECMWF forecasts of temperature, relative humidity, and wind.
• For the correction of the 10-m wind direction forecast, which is often difficult to achieve, CU-net also improves the correction performance.
Rights and permissions
Open AccessThis article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
About this article
Cite this article
Han, L., Chen, M., Chen, K.et al.A Deep Learning Method for Bias Correction of ECMWF 24–240 h Forecasts. Adv. Atmos. Sci.38,1444–1459 (2021). https://doi.org/10.1007/s00376-021-0215-y
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI:https://doi.org/10.1007/s00376-021-0215-y