NowDeepN: An Ensemble of Deep Learning Models for Weather Nowcasting Based on Radar Products’ Values Prediction
Source abstract
Abstract
One of the hottest topics in today’s meteorological research is Weather nowcasting, which is the weather forecast for a short time period such as one to six hours. Radar is an important data source used by operational meteorologists for issuing nowcasting warnings. With the main goal of helping meteorologists in analysing radar data for issuing nowcasting warnings, we propose NowDeepN, a supervised learning based regression model which uses an ensemble of deep artificial neural networks for predicting the values for radar products at a certain time moment. The values predicted by NowDeepN may be used by meteorologists in estimating the future development of potential severe phenomena and would replace the time consuming process of extrapolating the radar echoes. NowDeepN is intended to be a proof of concept for the effectiveness of learning from radar data relevant patterns that would be useful for predicting future values for radar products based on their historical values. For assessing the performance of NowDeepN, a set of experiments on real radar data provided by the Romanian National Meteorological Administration is conducted. The impact of a data cleaning step introduced for correcting the erroneous radar products’ values is investigated both from the computational and meteorological perspectives. The experimental results also indicate the relevance of the features considered in the supervised learning task, highlighting that the radar products’ values at a certain geographical location at a time moment may be predicted from the products’ values from a neighboring area of that location at previous time moments. An overall Normalized Root Mean Squared Error less than 4% was obtained for NowDeepN on the cleaned radar data. Compared to similar related work from the nowcasting literature, NowDeepN outperforms several approaches and this emphasizes the performance of our proposal.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Meteorological Phenomena and Simulations · Precipitation Measurement and Analysis · Climate variability and models
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Department of Computer Science, Babeş-Bolyai University, 400084 Cluj-Napoca, Romania
Romanian National Meteorological Administration, 013686 Bucharest, Romania
Funding
Norway Grants
References (55)
Show source-provided references
1. Franch, G., Nerini, D., Pendesini, M., Coviello, L., Jurman, G., and Furlanello, C. (2020). Precipitation Nowcasting with Orographic Enhanced Stacked Generalization: Improving Deep Learning Predictions on Extreme Events. Atmosphere, 11.10.3390/atmos11030267 ↗
2. A Deep Learning-Based Methodology for Precipitation Nowcasting With Radar10.1029/2019EA000812 ↗
3. Swedish Meteorological and Hydrological Institute (2018, July 10). Cooperation Is a Must for Adaptation to and Mitigation of Climate Change. Available online: https://www.smhi.se/en/news-archive/.
4. WMO—World Meteorological Organisation (2018, July 10). Weather Climate Water. Available online: https://www.wmo.int.
5. Numerical weather prediction10.1016/S0167-6105(02)00261-1 ↗
6. NOAA’s Radar Operations Center (2018, July 10). NEXRAD Technical Information, Available online: https://www.roc.noaa.gov/WSR88D/Engineering/NEXRADTechInfo.aspx.
7. Hering, A.M., Morel, C., Galli, G., Sénési, S., Ambrosetti, P., and Boscacci, M. (2014). Nowcasting thunderstorms in the Alpine region using a radar based adaptive thresholding scheme. Proceedings of ERAD, Copernicus GmbH.
8. James, P., Reichert, B., and Heizenreder, D. (2015). NowCastMIX—optimized automatic warnings from continuously monitored nowcasting systems based on fuzzy-logic evaluations of storm attributes. European Conference on Severe Storms, American Meteorological Society.
9. AROME-NWC Overview, Results, Evolution and Perspectives
10. TITAN: Thunderstorm Identification, Tracking, Analysis, and Nowcasting—A radar-based methodology10.1175/1520-0426(1993)010<0785:TTITAA>2.0.CO;2 ↗
11. The Storm Cell Identification and Tracking Algorithm: An Enhanced WSR-88D Algorithm10.1175/1520-0434(1998)013<0263:TSCIAT>2.0.CO;2 ↗
12. Radar-based cell tracking with fuzzy logic approach10.1002/met.1509 ↗
13. AROME—NWC: A new nowcasting tool based on an operational mesoscale forecasting system10.1002/qj.2463 ↗
14. The Integrated Nowcasting through Comprehensive Analysis (INCA) System and Its Validation over the Eastern Alpine Region10.1175/2010WAF2222451.1 ↗
15. IBM (2019, September 15). Deep Thunder. Available online: http://www-03.ibm.com/ibm/history/ibm100/us/en/icons/deepthunder/.
16. Panasonic (2019, September 15). Panasonic Global 4D Weather. Available online: https://www-media.panasonic.aero/2016/01/010616-PWS4DGlobal.pdf.
17. Climacell (2019, September 15). The ClimaCell Engine. Available online: https://www.climacell.co/.
18. TempoQuest (2019, September 15). TempoQuest. Available online: http://tempoquest.com/.
19. Mitchell, T.M. (1997). Machine Learning, McGraw-Hill, Inc.. [1st ed.].
20. Goodfellow, I., Bengio, Y., and Courville, A. (2016). Deep Learning, MIT Press.
21. Dropout: A simple way to prevent neural networks from overfitting
22. Forecasting bank failures and stress testing: A machine learning approach10.1016/j.ijforecast.2018.01.009 ↗
23. Predicting the failures of prediction markets: A procedure of decision making using classification models10.1016/j.ijforecast.2018.04.003 ↗
24. Wage against the machine: A generalized deep-learning market test of dataset value10.1016/j.ijforecast.2017.09.008 ↗
25. Simplifying Neural Networks by Soft Weight-Sharing10.1162/neco.1992.4.4.473 ↗
26. Le, Q. (2013, January 26–31). Building high-level features using large scale unsupervised learning. Proceedings of the 2013 IEEE International Conference on Acoustics, Speech and Signal Processing, Vancouver, BC, Canada.10.1109/ICASSP.2013.6639343 ↗
27. Deng, J., Zhang, Z., Marchi, E., and Schuller, B. (2013, January 2–5). Sparse autoencoder-based feature transfer learning for speech emotion recognition. Proceedings of the Humaine Association Conference on Affective Computing and Intelligent Interaction, Geneva, Switzerland.10.1109/ACII.2013.90 ↗
28. Teletin, M., Czibula, G., Bocicor, M.I., Albert, S., and Pandini, A. (2018). Deep Autoencoders for Additional Insight into Protein Dynamics. International Conference on Artificial Neural Networks, Springer.10.1007/978-3-030-01421-6_8 ↗
29. Teletin, M., Czibula, G., and Codre, C. (2019). AutoSimP: An Approach for Predicting Proteins’ Structural Similarities Using an Ensemble of Deep Autoencoders. International Conference on Knowledge Science, Engineering and Management, Springer.10.1007/978-3-030-29563-9_5 ↗
30. Principal component analysis: A review and recent developments10.1098/rsta.2015.0202 ↗
31. A machine learning nowcasting method based on real-time reanalysis data10.1002/2016JD025783 ↗
32. Beusch, L., Clementi, L., Foresti, L., Hamann, U., Hering, A.M., Leonarduzzi, E., Nerini, D., Nisi, L., Sassi, M., and Germann, U. (2017, January 18–22). Towards thunderstorm nowcasting by applying machine learning to a multi-sensor observation and NWP model database. Proceedings of the European Conference on Severe Storms 2017, Pula, Croatia.
33. Shi, X., Chen, Z., Wang, H., Yeung, D.Y., Wong, W.k., and Woo, W.c. Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting. Proceedings of the 28th International Conference on Neural Information Processing Systems—Volume 1.
34. Kim, S., Hong, S., Joh, M., and Song, S. (2017). DeepRain: ConvLSTM Network for Precipitation Prediction using Multichannel Radar Data. arXiv.
35. Heye, A., Venkatesan, K., and Cain, J. (2017). Precipitation Nowcasting: Leveraging Deep Convolutional Recurrent Neural Networks. Proceedings of the Cray User Group (CUG), American Meteorological Society.
36. Deep Learning for Precipitation Nowcasting: A Benchmark and A New Model
37. Meteonowcasting using Deep Learning Architecture
38. Nowcasting Foehn Wind Events Using the AdaBoost Machine Learning Algorithm10.1175/WAF-D-16-0208.1 ↗
39. Ji, Y. (2017, January 26–27). Short-term Precipitation Prediction Based on a Neural Network. Proceedings of the 3rd International Conference on Artificial Intelligence and Industrial Engineering, AIIE 2017, Shanghai, China.
40. Tran, Q.K., and Song, S.k. (2019). Computer Vision in Precipitation Nowcasting: Applying Image Quality Assessment Metrics for Training Deep Neural Networks. Atmosphere, 10.10.3390/atmos10050244 ↗
41. Convolutional Neural Network for Convective Storm Nowcasting Using 3D Doppler Weather Radar Data10.1109/TGRS.2019.2948070 ↗
42. Convolutional Residual-Attention: A Deep Learning Approach for Precipitation Nowcasting
43. An Experimental Comparison of Three Methods for Constructing Ensembles of Decision Trees: Bagging, Boosting, and Randomization10.1023/A:1007607513941 ↗
44. Nowcasting: The Promise of New Technologies of Communication, Modeling, and Observation10.1175/BAMS-D-11-00153.1 ↗
45. Using self-organizing maps for unsupervised analysis of radar data for nowcasting purposes10.1016/j.procs.2019.09.159 ↗
46. Agarap, A.F. (2018). Deep Learning using Rectified Linear Units (ReLU). arXiv.
47. Kingma, D., and Ba, J. (2014). Adam: A method for stochastic optimization. arXiv.
48. Another look at measures of forecast accuracy10.1016/j.ijforecast.2006.03.001 ↗
49. Interval Estimation for a proportion10.1214/ss/1009213286 ↗
50. Keras (2020, December 15). The Python Deep Learning Library. Available online: https://keras.io/.
51. Mihai, A. (2020, December 15). NowDeepN Software. Available online: https://github.com/mihaiandrei1294/NowDeepN.
52. Scikit-learn (2020, December 15). Machine Learning in Python. Available online: http://scikit-learn.org/stable/.
53. Convolutional neural networks: An overview and application in radiology10.1007/s13244-018-0639-9 ↗
54. A novel approach for software defect prediction through hybridizing gradual relational association rules with artificial neural networks10.1016/j.ins.2018.02.027 ↗
55. Incremental relational association rule mining of educational data sets