Multi-Step Prediction of Typhoon Tracks Combining Reanalysis Image Fusion Using Laplacian Pyramid and Discrete Wavelet Transform with ConvLSTM
<p>Typhoon moving angle diagram.</p> "> Figure 2
<p>The training process of improved ConvLSTM.</p> "> Figure 3
<p>ConvLSTM unit structure.</p> "> Figure 4
<p>The diagram of adding weight matrixes to memory cells in gate unit calculation.</p> "> Figure 5
<p>The process of establishing Laplacian Pyramid.</p> "> Figure 6
<p>The diagram of multi-step prediction.</p> "> Figure 7
<p>Reanalysis images of physical variables group with marked points (<b>a</b>) Mean wave direction (<b>b</b>) Mean wave period (<b>c</b>) Significant height of combined wind waves and swell (<b>d</b>) 10 m v-component of wind.</p> "> Figure 8
<p>The values of weights of memory cells at last two moments in each series (<b>a</b>) and the results diagram of single-step prediction (<b>b</b>).</p> "> Figure 9
<p>The comparison of predicted images and real images at two steps in 12 h prediction of Typhoon IN-FA (202106) (at 18 o’clock on 22 July 2021 and at 0 o’clock on 23 July 2021) ((<b>a</b>) Predicted images at first step; (<b>c</b>) Real images at first step; (<b>b</b>) Predicted images at second step; (<b>d</b>) Real images at second step) and Typhoon RAI (202122) (at 6 o’clock on 19 December 2021 and at 12 o’clock on 19 December 2021) ((<b>e</b>) Predicted images at first step; (<b>g</b>) Real images at first step; (<b>f</b>) Predicted images at second step; (<b>h</b>) Real images at second step).</p> "> Figure 10
<p>The diagram of comparisons between predicted tracks and real tracks in 12 h (<b>a</b>), 18 h (<b>b</b>), 24 h (<b>c</b>) and 48 h (<b>d</b>) of typhoon CEMPAKA (202107).</p> "> Figure 11
<p>The diagram of comparisons between predicted tracks and real tracks in 12 h (<b>a</b>), 18 h (<b>b</b>), 24 h (<b>c</b>) and 48 h (<b>d</b>) of typhoon KOMPASU (202118).</p> ">
Abstract
:1. Introduction
2. Materials and Methods
2.1. Related Work
2.1.1. ConvLSTM
2.1.2. Spatial Attention
- Firstly, and are obtained by operating maximum pooling and average pooling to input feature along the channel direction, as shown in Formula (6) and (7).
- Then, , namely, the weight matrix, is obtained by cascading and , the convolution operation and the Sigmoid activation function operation, as shown in Formula (8),
- Finally, the output features are calculated by Formula (9).
2.1.3. Laplacian Pyramid
2.1.4. Discrete Wavelet Transform
2.2. ConvLSTM Establishment and Prediction
2.2.1. Correlation Analysis of Reanalysis Data
2.2.2. The Establishment and Optimization of ConvLSTM
2.2.3. Fusion with Similar Images
2.2.4. Multi-Step Prediction of Typhoon Tracks
2.3. Datasets Establishment
2.3.1. Data of Typhoon Center Coordinates
2.3.2. Reanalysis Data
2.3.3. Data Processing
3. Results
3.1. Experimental Environment
3.2. Model Evaluation Index
3.3. Experimental Results
3.4. Comparison with Results of Other Methods
4. Discussion
5. Conclusions
Author Contributions
Funding
Data Availability Statement
Conflicts of Interest
References
- Yuan, S.J.; Wang, C.; Mu, B.; Zhou, F.F.; Duan, W.S. Typhoon Intensity Forecasting Based on LSTM Using the Rolling Forecast Method. Algorithms 2021, 14, 83. [Google Scholar] [CrossRef]
- Heidarzadeh, M.; Iwamoto, T.; Takagawa, T.; Takagi, H. Field surveys and numerical modeling of the August 2016 Typhoon Lionrock along the northeastern coast of Japan: The first typhoon making landfall in Tohoku region. Nat. Hazards 2021, 105, 1–19. [Google Scholar] [CrossRef]
- Eadie, P. Typhoon Yolanda and post-disaster resilience: Problems and challenges. Asia Pac. Viewp. 2019, 60, 94–107. [Google Scholar] [CrossRef]
- Shimozono, T.; Tajima, Y.; Kumagai, K.; Arikawa, T.; Oda, Y.; Shigihara, Y.; Mori, N.; Suzuki, T. Coastal impacts of super typhoon Hagibis on Greater Tokyo and Shizuoka areas, Japan. Coast. Eng. J. 2020, 62, 129–145. [Google Scholar] [CrossRef]
- Hon, K. Tropical cyclone track prediction using a large-area WRF model at the Hong Kong Observatory. Trop. Cyclone Res. Rev. 2020, 9, 67–74. [Google Scholar] [CrossRef]
- Lian, J.; Dong, P.P.; Zhang, Y.P.; Pan, J.G. A Novel Deep Learning Approach for Tropical Cyclone Track Prediction Based on Auto-Encoder and Gated Recurrent Unit Networks. Appl. Sci. 2020, 10, 3965. [Google Scholar] [CrossRef]
- Tong, B.; Wang, X.; Fu, J.Y.; Chan, P.; He, Y. Short-term prediction of the intensity and track of tropical cyclone via ConvLSTM model. J. Wind Eng. Ind. Aerodyn. 2022, 226, 105026. [Google Scholar] [CrossRef]
- Yasunaga, K.; Miyajima, T.; Yamaguchi, M. Relationships between Tropical Cyclone Motion and Surrounding Flow with Reference to Longest Radius and Maximum Sustained Wind. Sola 2016, 12, 277–281. [Google Scholar] [CrossRef]
- Ruttgers, M.; Lee, S.; Jeon, S.; You, D. Prediction of a typhoon track using a generative adversarial network and satellite images. Sci. Rep. 2019, 9, 6057. [Google Scholar] [CrossRef]
- Mei, X.G.; Pan, e.; Ma, Y.; Dai, X.B.; Huang, J.; Fan, F.; Du, Q.D.; Zheng, H.; Ma, J.Y. Spectral–Spatial Attention Network for Hyperspectral Image Classification. Remote Sens. 2019, 11, 963. [Google Scholar] [CrossRef]
- Chandra, R.; Goyal, S.; Gupta, R. Evaluation of Deep Learning Models for Multi-Step Ahead Time Series Prediction. IEEE Access 2021, 9, 83105–83123. [Google Scholar] [CrossRef]
- Chang, C.Y.; Lu, C.W.; Wang, C.A.J. A Multi-Step-Ahead Markov Conditional Forward Model with Cube Perturbations for Extreme Weather Forecasting. In Proceedings of the 35th AAAI Conference on Artificial Intelligence, Online, 2–9 February 2021. [Google Scholar]
- Deo, R.; Chandra, R. Multi-Step-Ahead Cyclone Intensity Prediction with Bayesian Neural Networks. In Proceedings of the 16th Pacific Rim International Conference on Artificial Intelligence (PRICAI), Cuvu, Fiji, 26–30 August 2019. [Google Scholar]
- Bing, Q.C.; Shen, F.X.; Chen, X.F.; Zhang, W.J.; Hu, Y.R.; Qu, D.Y. A Hybrid Short-Term Traffic Flow Multistep Prediction Method Based on Variational Mode Decomposition and Long Short-Term Memory Model. Discret. Dyn. Nat. Soc. 2021, 2021, 4097149. [Google Scholar] [CrossRef]
- Du, Z.H.; Qin, M.J.; Zhang, F.; Liu, R.Y. Multistep-ahead forecasting of chlorophyll a using a wavelet nonlinear autoregressive network. Knowl.-Based Syst. 2018, 160, 61–70. [Google Scholar] [CrossRef]
- Zhao, F.; Gao, Y.T.; Li, X.N.; An, Z.Y.; Ge, S.Y.; Zhang, C.M. A similarity measurement for time series and its application to the stock market. Expert Syst. Appl. 2021, 182, 115217. [Google Scholar] [CrossRef]
- Mao, R.; Fu, X.S.; Niu, P.J.; Wang, H.Q.; Pan, J.; Li, S.S.; Liu, L. Multi-directional Laplacian Pyramid Image Fusion Algorithm. In Proceedings of the 3rd International Conference on Mechanical, Control and Computer Engineering (ICMCCE), Huhhot, China, 14–16 September 2018. [Google Scholar]
- Sharma, A.; Gupta, H.; Sharma, Y. Image Fusion with Deep Leaning using Wavelet Transformation. J. Emerg. Technol. Innov. Res. 2021, 8, 2826–2834. [Google Scholar]
- Shi, X.J.; Chen, Z.R.; Wang, H.; Yeung, D.Y.; Wong, W.K.; Woo, W.C. Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting. In Proceedings of the 29th Annual Conference on Neural Information Processing Systems (NIPS), Montreal, QB, Canada, 7–12 December 2015. [Google Scholar]
- Moishin, M.; Deo, R.C.; Prasad, R.; Raj, N.; Abdulla, S. Designing Deep-Based Learning Flood Forecast Model With ConvLSTM Hybrid Algorithm. IEEE Access 2021, 9, 50982–50993. [Google Scholar] [CrossRef]
- Woo, S.H.; Park, J.; Lee, J.Y.; Kweon, I.S. CBAM: Convolutional Block Attention Module. In Proceedings of the 15th European Conference on Computer Vision (ECCV), Munich, Germany, 8–14 September 2018. [Google Scholar]
- Guo, C.L.; Szemenyei, M.; Fan, C.Q. SA-UNet: Spatial Attention U-Net for Retinal Vessel Segmentation. In Proceedings of the 25th International Conference on Pattern Recognition (ICPR), Online, 10–15 January 2021. [Google Scholar]
- Liu, F.Q.; Chen, L.H.; Lu, L.; Ahmad, A.; Jeon, G.; Yang, X.M. Medical image fusion method by using Laplacian pyramid and convolutional sparse representation. Concurr. Comput.-Pract. Exp. 2020, 32, e5632. [Google Scholar] [CrossRef]
- Wang, J.; Ke, C.; Wu, M.H.; Liu, M.; Zeng, C.Y. Infrared and visible image fusion based on Laplacian pyramid and generative adversarial network. KSII Trans. Internet Inf. Syst. 2021, 15, 1761–1777. [Google Scholar]
- Starosolski, R. Hybrid Adaptive Lossless Image Compression Based on Discrete Wavelet Transform. Entropy 2020, 22, 751. [Google Scholar] [CrossRef]
- Kambalimath, S.S.; Deka, P.C. Performance enhancement of SVM model using discrete wavelet transform for daily streamflow forecasting. Environ. Earth Sci. 2021, 80, 101. [Google Scholar] [CrossRef]
- Kanagaraj, H.; Muneeswaran, V. Image Compression Using HAAR Discrete Wavelet Transform. In Proceedings of the 5th International Conference on Devices, Circuits and Systems (ICDCS), Coimbatore, India, 5–6 March 2020. [Google Scholar]
- Mahmoud, H.; Akkari, N. Shortest Path Calculation: A Comparative Study for Location-Based Recommender System. In Proceedings of the World Symposium on Computer Applications and Research (WSCAR), Cairo, Egypt, 12–14 March 2016. [Google Scholar]
- Kitamoto, A. “Digital Typhoon” Typhoon Analysis Based on Artificial Intelligence Approach; Technical Report of Information Processing Society of Japan (IPSJ); CVIM123-8; Processing Society of Japan (IPSJ): Tokyo, Japan, 2000; pp. 59–66. [Google Scholar]
- Hersbach, H.; Bell, B.; Berrisford, P.; Hirahara, S.; Horányi, A.; Muñoz-Sabater, J.; Nicolas, J.; Peubey, C.; Radu, R.; Schepers, D.; et al. The ERA5 global reanalysis. Q. J. R. Meteorol. Soc. 2020, 146, 1999–2049. [Google Scholar] [CrossRef]
The Difference of Latitude Coordinates of Typhoon Centers at Current and Previous Moments ≥0° | The Difference of Latitude Coordinates of Typhoon Centers at Current and Previous Moments <0° | Total Typhoon Moments | |
---|---|---|---|
Latitude coordinates of the typhoon center at the current moment < 15° | 706 | 187 | 893 |
Latitude coordinates of the typhoon center at the current moment > 15° | 3559 | 415 | 3974 |
Latitude coordinates of the typhoon center at the current moment = 15° | 7 | 4 | 11 |
Total typhoon moments | 4272 | 606 | 4878 |
The Difference of Longitude Coordinates of Typhoon Centers at Current and Previous Moments >0° | The Difference of Longitude Coordinates of Typhoon Centers at Current and Previous Moments ≤0° | Total Typhoon Moments | |
---|---|---|---|
Latitude coordinates of the typhoon center at the current moment < 15° | 32 | 861 | 893 |
Latitude coordinates of the typhoon center at the current moment > 15° | 1713 | 2261 | 3974 |
Latitude coordinates of the typhoon center at the current moment = 15° | 0 | 11 | 11 |
Total typhoon moments | 1745 | 3133 | 4878 |
Parameters and the Name of Equipment | Version |
---|---|
GPU | NVIDIA GeForce RTX 3060 |
epoch | 50 |
batch_size | 10 |
learning rate | 0.001 |
Parameters | Values |
---|---|
Input dimension | (batch_size, 12, 64, 64) |
Kernel sizes | (3 × 3) |
Number of Gate units nodes | 12 |
Number of layers | 3 |
Physical Variables | The Correlation with Typhoon Moving Angle and Distance | Weights |
---|---|---|
Mean wave direction | 0.6827636635668061 | 0.45 |
Mean wave period | 0.6709704103073449 | 0.15 |
Significant height of combined wind waves and swell | 0.6650877022571118 | 0.1 |
10 m v-component of wind | 0.6774975957516751 | 0.3 |
Methods | MAE | RMSE |
---|---|---|
Model1 | 131.88 | 186.59 |
Model2 | 103.79 | 130.80 |
Model3 | 102.14 | 127.77 |
Models | Predicted Time | MAE (km) | RMSE (km) |
---|---|---|---|
Proposed method in this paper | 12 h | 102.14 | 127.77 |
18 h | 168.17 | 209.19 | |
24 h | 243.73 | 300.67 | |
48 h | 574.62 | 694.99 | |
LSTM | 12 h | 245.78 | 358.56 |
18 h | 318.02 | 433.79 | |
24 h | 488.36 | 678.79 | |
48 h | 1128.14 | 1555.05 | |
GRU | 12 h | 358.38 | 511.23 |
18 h | 330.41 | 450.35 | |
24 h | 376.63 | 759.32 | |
48 h | 575.67 | 704.55 | |
Model 2 in Section 3.3 | 12 h | 103.79 | 130.80 |
18 h | 174.10 | 225.32 | |
24 h | 248.33 | 314.62 | |
48 h | 587.71 | 705.22 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Lu, P.; Xu, M.; Chen, M.; Wang, Z.; Zheng, Z.; Yin, Y. Multi-Step Prediction of Typhoon Tracks Combining Reanalysis Image Fusion Using Laplacian Pyramid and Discrete Wavelet Transform with ConvLSTM. Axioms 2023, 12, 874. https://doi.org/10.3390/axioms12090874
Lu P, Xu M, Chen M, Wang Z, Zheng Z, Yin Y. Multi-Step Prediction of Typhoon Tracks Combining Reanalysis Image Fusion Using Laplacian Pyramid and Discrete Wavelet Transform with ConvLSTM. Axioms. 2023; 12(9):874. https://doi.org/10.3390/axioms12090874
Chicago/Turabian StyleLu, Peng, Mingyu Xu, Ming Chen, Zhenhua Wang, Zongsheng Zheng, and Yixuan Yin. 2023. "Multi-Step Prediction of Typhoon Tracks Combining Reanalysis Image Fusion Using Laplacian Pyramid and Discrete Wavelet Transform with ConvLSTM" Axioms 12, no. 9: 874. https://doi.org/10.3390/axioms12090874