Abstract
Reliable precipitation nowcasting is essential to many fields, which can guide people to reasonably carry out production activities and respond to rainstorm disasters. However, precipitation nowcasting is a very challenging task because of correlation and heterogeneity both in space and in time. Most previous studies have not adequately captured the long-term and long-range spatiotemporal dependencies in the data, leading to insufficient modeling and poor prediction performance. To make more accurate prediction, we propose a novel deep learning model for precipitation nowcasting, called two-stream convolutional LSTM which includes short-term sub-network and long-term sub-network. The two sub-networks, respectively, make predictions on inputs at different time intervals to capture the heterogeneity of rainfall data. On this basis, an innovative recombination module is proposed to fuse the outputs of two sub-networks. In addition, we embed the 3D convolutions and self-attention mechanism to construct a new memory cell, named 3D-SA-LSTM, to extract the spatiotemporal feature. Two-stream convolutional LSTM achieves the state-of-the-art prediction performance on a real-world large-scale dataset and is a more flexible framework that can be conveniently applied to other similarly time series prediction tasks: traffic forecasting and planning, financial analysis and management, actions recognition and prediction, etc.
Similar content being viewed by others
References
Yuehong S, Wanchang Z, Yonghe L (2009) Application of back-propagation neural network in precipitation estimation with doppler radar. Plateau Meteor 28(4):846–853
Heuvelink D, Berenguer M, Brauer CC et al (2020) Hydrological application of radar rainfall nowcasting in the Netherlands. Environ Int 136:105431. https://doi.org/10.1016/j.envint.2019.105431
Sun J, Xue M, Wilson JW et al (2014) Use of NWP for nowcasting convective precipitation: Recent progress and challenges. Bull Am Meteor Soc 95(3):409–426. https://doi.org/10.1175/BAMS-D-11-00263.1
Wang G, Wong W, Liu L et al (2013) Application of multi-scale tracking radar echoes scheme in quantitative precipitation nowcasting. Adv Atmos Sci 30(2):448–460. https://doi.org/10.1175/BAMS-D-11-00263.1
Li H, Wang X, Wu S et al (2021) A new method for determining an optimal diurnal threshold of GNSS precipitable water vapor for precipitation forecasting. Remote Sens 13(7):1390. https://doi.org/10.3390/rs13071390
Chen M, Bica B, Tüchler L et al (2017) Statistically extrapolated nowcasting of summertime precipitation over the Eastern Alps. Adv Atmos Sci 34:925–938. https://doi.org/10.1007/s00376-017-6185-4
LeCun Y, Bengio Y, Hinton G (2015) Deep learning. Nature 521(7553):436–444. https://doi.org/10.1038/nature14539
Shi X, Chen Z, Wang H, Yeung DY, Wong WK, Woo WC (2015) Convolutional LSTM network: a machine learning approach for precipitation nowcasting. In: Advances in neural information processing systems, p 802–810.
Kong Y, Fu Y (2018) Human action recognition and prediction: A survey. arXiv:1806.11230
Sahin C, Garcia-Hernando G, Sock J et al (2020) A review on object pose recovery: from 3d bounding box detectors to full 6d pose estimators. Image Vis Comput 96:103898. https://doi.org/10.1016/j.imavis.2020.103898
Sun H, Tang M, Peng W et al (2021) Interval prediction of short-term building electrical load via a novel multi-objective optimized distributed fuzzy model. Neural Comput & Applic. https://doi.org/10.1007/s00521-021-06162-9
Boogaard MV, Pickkers P, Hoeven HV et al (2010) PREDICT, Prediction of Delirium in ICU Patients: development and validation of a prediction model. Crit Care 14:P498. https://doi.org/10.1186/cc8730
Chen Z, Xiao X, Li C et al (2016) Erratum to: Real-time transient stability status prediction using cost-sensitive extreme learning machine. Neural Comput & Applic 27:333. https://doi.org/10.1007/s00521-015-1926-8
Baehr J, Fröhlich K, Botzet M et al (2015) The prediction of surface temperature in the new seasonal prediction system based on the MPI-ESM coupled climate model. Clim Dyn 44(9–10):2723–2735. https://doi.org/10.1007/s00382-014-2399-7
Kosek W, McCarthy DD, Luzum BJ (2001) El Niño impact on polar motion prediction errors. Stud Geophys Geod 45(4):347–361. https://doi.org/10.1023/A:1022073503034
Chiu PC, Selamat A, Krejcar O et al (2021) Imputation of rainfall data using the sine cosine function fitting neural network. Int J Interact Multimed and Art Intellig. https://doi.org/10.9781/ijimai.2021.08.013
Sutskever I, Vinyals O, Le Q V (2014) Sequence to sequence learning with neural networks. arXiv:1409.3215.
Schmidhuber J (1992) Learning complex, extended sequences using the principle of history compression. Neural Comput 4(2):234–242. https://doi.org/10.1162/neco.1997.9.8.1735
Cho, K., van Merriënboer, B., Gulcehre, C., et al. (2014). Learning phrase representations using RNN encoder-decoder for statistical machine translation. arXiv:1406.1078
Srivastava N, Mansimov E, Salakhutdinov R (2015) Unsupervised learning of video representations using LSTMS. In: International conference on machine learning, p 843–852
Ranzato MA, Huang FJ, Boureau YL et al (2007) Unsupervised learning of invariant feature hierarchies with applications to object recognition[C]. IEEE Conf Comput Vis Pattern Recogn 2007:1–8
Finn C, Goodfellow I, Levine S (2016) Unsupervised learning for physical interaction through video prediction. arXiv:1605.07157.
Kalchbrenner N, van den Oord A, Simonyan K, Danihelka I, Vinyals O, Graves A, Kavukcuoglu K (2017) Video pixel networks. In: Proceedings of the international conference on machine learning. p 1771–1779
Vondrick C, Pirsiavash H, Torralba A (2016) Generating videos with scene dynamics. arXiv:1609.02612
Mathieu M, Couprie C, LeCun Y (2015) Deep multi-scale video prediction beyond mean square error. arXiv:1511.05440
Goodfellow IJ, Pouget-Abadie J, Mirza M, Xu B, Warde-Farley D, Ozairy S, Courville A, Bengio Y (2014) Generative adversarial nets. arXiv:1406.2661v1
Tran D, Bourdev L, Fergus R, Torresani L, Paluri M (2015) Learning spatiotemporal features with 3d convolutional networks. In: Proceedings of the IEEE International conference on computer vision, p 4489–4497. https://doi.org/10.1109/ICCV.2015.510.
Ji S, Xu W, Yang M, Yu K (2013) 3D convolutional neural networks for human action recognition. IEEE TransPattern Anal Mach Intell 35(1):221–231. https://doi.org/10.1109/TPAMI.2012.59
Wang X, Girshick R, Gupta A, He K (2018) Non-local neural networks. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, p 7794–7803. https://doi.org/10.1109/CVPR.2018.00813
Vaswani A, Shazeer N, Parmar N et al (2017) Attention is all you need. arXiv:1706.03762
Huddar MG, Sannakki SS, Rajpurohit VS (2021) Attention-based multi-modal sentiment analysis and emotion detection in conversation using RNN. Int J Inter Multimed and Art Intellig. https://doi.org/10.9781/ijimai.2020.07.004
Klein B, Wolf L, Afek Y (2015) A dynamic convolutional layer for short range weather prediction. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, p 4840–4848. https://doi.org/10.1109/CVPR.2015.7299117
Woo W, Wong W (2017) Operational application of optical flow techniques to radar-based rainfall nowcasting. Atmos 8:48. https://doi.org/10.3390/ATMOS8030048
Shi X, Gao Z, Lausen L et al (2017) Deep learning for precipitation nowcasting: a benchmark and a new model. arXiv:1706.03458
WangY, Long M, Wang J, Gao Z, Yu P S (2017) PredRNN: recurrent neural networks for predictive learning using spatiotemporal LSTMs. In: Proceedings of the Neural Information Processing Systems, p 879–888.
Wang Y, Jiang L, Yang M H, Li L J, Long M, Li F (2019) Eidetic 3d lstm: A model for video prediction and beyond. In: Proceedings of the International Conference on Learning Representations
Wang Y, Zhang J, Zhu H, Long M, Wang J, Yu P S (2019) Memory in memory: a predictive neural network for learning higher-order non-stationarity from spatiotemporal dynamics. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, p 9154–9162. https://doi.org/10.1109/CVPR.2019.00937.
Guen, V L, Thome N (2020) Disentangling physical dynamics from unknown factors for unsupervised video prediction. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, p. 11474–11484. https://doi.org/10.1109/CVPR42600.2020.01149
Luo W, Li Y, Urtasun R, Zemel R (2016) Understanding the effective receptive field in deep convolutional neural networks. In: Proceedings of the 30th International Conference on Neural Information Processing Systems (NIPS), p 4905–4913
Hu J, Shen L, Sun G (2018) Squeeze-and-excitation networks. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, p 7132–7141. https://doi.org/10.1109/TPAMI.2019.2913372
Simonyan K, Zisserman A (2014) Two-stream convolutional networks for action recognition in videos. arXiv:1406.2199
He K, Zhang X, Ren S et al (2016) Deep residual learning for image recognition, In: Proceedings of the 2016 IEEE conference on computer vision and pattern recognition, p 770–778. https://doi.org/10.1109/cvpr.2016.90
Kingma D P, Ba J (2014) Adam: A method for stochastic optimization. arXiv:1412.6980
Ba J L, Kiros J R, Hinton G E (2016) Layer normalization. arXiv:1607.06450
Hogan RJ, Ferro CAT, Jolliffe IT et al (2010) Equitability revisited: why the “equitable threat score” is not equitable. Weather Forecast 25:710–726. https://doi.org/10.1175/2009WAF2222350.1
Funding
This work was supported by the National Natural Science Foundation of China (Nos. 61906097) and a project funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD).
Author information
Authors and Affiliations
Contributions
All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Suting Chen and Xin Xu. The first draft of the manuscript was written by Xin Xu, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Corresponding author
Ethics declarations
Conflicts of interest
The authors declare that they have no conflict of interests.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
About this article
Cite this article
Chen, S., Xu, X., Zhang, Y. et al. Two-stream convolutional LSTM for precipitation nowcasting. Neural Comput & Applic 34, 13281–13290 (2022). https://doi.org/10.1007/s00521-021-06877-9
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-021-06877-9