Précédent|Prochain

Rapid correction of near real-time FY-4A retrieval based on ensemble machine learning

  • role： First author第一作者
  • Affiliation：

    National Key Laboratory of Water Disaster Prevention, Hohai University, Nanjing 210098, China

    College of Hydrology and Water Resources, Hohai University, Nanjing 210098, China

  • Email：lyuyi@hhu.edu.cn
  • Introduction：吕毅，研究方向为水文遥感和机器学习。E-mail：lyuyi@hhu.edu.cn
LYU Yi12,  
  • role： Corresponding author通信作者
  • Affiliation：

    National Key Laboratory of Water Disaster Prevention, Hohai University, Nanjing 210098, China

    College of Hydrology and Water Resources, Hohai University, Nanjing 210098, China

  • Email：yongbin@hhu.edu.cn
  • Introduction：雍斌，研究方向为遥感水文学。E-mail：yongbin@hhu.edu.cn
YONG Bin12*,  
  • Affiliation：

    National Key Laboratory of Water Disaster Prevention, Hohai University, Nanjing 210098, China

SHEN Zhehui1,  
  • Affiliation：

    National Key Laboratory of Water Disaster Prevention, Hohai University, Nanjing 210098, China

    College of Hydrology and Water Resources, Hohai University, Nanjing 210098, China

LI Ji12,  
  • Affiliation：

    National Key Laboratory of Water Disaster Prevention, Hohai University, Nanjing 210098, China

    College of Hydrology and Water Resources, Hohai University, Nanjing 210098, China

MEI Jun12

résumé

Satellite remote sensing retrieval is an important way to solve the problem of obtaining near-real-time high-resolution precipitation information. Fengyun-4A (FY-4A) is outfitted with the Advanced Geosynchronous Radiation Imager (AGRI), which boasts world-leading performance. The dual scanning mirrors of AGRI enable precise 2-D pointing, allowing for minute-level regional scans—a groundbreaking achievement. This advanced instrument can capture high-frequency images of the Earth’s cloud cover in more than 14 spectral bands. It can generate the official FY-4A REGC (Regional Precipitation Estimation Near-real-time Product for China), which is one of the precipitation estimates information that China can independently obtain from satellite remote retrieval. However, the accuracy of FY-4A REGC still lags behind that of IMERG-Early, the counterpart product of the Global Precipitation Measurement (GPM). Currently, the prevailing approach for correcting satellite-derived precipitation products involves constructing linear prior relationship models between historical satellite rainfall estimates and corresponding ground truth measurements, typically obtained from rain gauges or radar systems. When new observational data become available, this relationship is utilized to derive corrected precipitation values. However, linear models struggle to precisely capture the intricate relationship between satellite rainfall estimates and ground truth measurements. We have observed that ensemble learning methods offer nonlinear models that exhibit advantages such as faster training, reduced data requirements, and robust model stability. In this study, a correction method for official FY-4A precipitation estimates is dynamically constructed using an ensemble machine learning method (LightGBM) with FY-4A REGC as the model input and IMERG-Early as the training calibration for the mainland China region. The revised FY-4A precipitation product (FY-4A Adj) was compared with the original FY-4A REGC using the CMPA automatic gauge observations as the ground reference. The Correlation Coefficient (CC), Root Mean Square Error (RMSE), and relative bias (Bias) of FY-4A Adj were found to be improved significantly compared with those of FY-4A REGC. The revised algorithm effectively reduced the significant overestimation of the original FY-4A REGC in southern China. Our investigation revealed that choosing the correct order for training information significantly enhances model accuracy, with this study opting for training order 221. In practical applications, the ensemlde learning model can continually optimize its model parameters and performance by dynamically adjusting to the latest training data in real time. We also conducted a comparative analysis of two classes of methods employing ensemble learning, namely, bagging and boosting. Our findings indicate that the Random Forest method performs better when working with limited data volumes, while LightGBM is the recommended choice for large datasets. In conclusion, the correction method based on ensemble machine learning proposed in this paper can quickly and effectively improve the near-real-time Precipitation estimates of FY-4A REGC. This method provides guidance for producing high-quality satellite precipitation products based on FY-4A.

mots-clés

remote sensing;Fengyun satellite;FY-4A;ensemble learning;near real-time precipitation estimates;precipitation estimates correction

References

  1. 1.
    Bottou L and Bousquet O. 2007. The tradeoffs of large scale learning//Proceedings of the 20th International Conference on Neural Information Processing Systems. Vancouver: Curran Associates Inc.: 161-168
  2. 2.
    Breiman L. 2001. Random forest. Machine Learning, 45(1): 5-32
  3. 3.
    Chauhan N K and Singh K. 2018. A review on conventional machine learning vs deep learning//2018 International Conference on Computing, Power and Communication Technologies (GUCON). Greater Noida: IEEE: 347-352
  4. 4.
    Chen H, Ning C, Nan Z T, Wang Y D, Wu X B and Zhao L. 2017. Correction of the daily precipitation data over the Tibetan Plateau with machine learning models. Journal of Glaciology and Geocryology, 39(3): 583-592
  5. 5.
    Chen K and Zhu Y. 2007. A summary of machine learning and related algorithms. Statistics and Information Forum, 22(5): 105-112
  6. 6.
    Chen T Q and Guestrin C. 2016. XGBoost: a scalable tree boosting system//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Francisco: ACM: 785-794
  7. 7.
    Chen X H, Zhong R D, Wang Z L, Lai C G and Chen J C. 2017. Evaluation on the accuracy and hydrological performance of the latest-generation GPM IMERG product over South China. Journal of Hydraulic Engineering, 48(10): 1147-1156
  8. 8.
    Dong Y H. 2016. FY-4 meteorological satellite and its application prospect. Aerospace Shanghai, 33(2): 1-8
  9. 9.
    Draper D W, Newell D A, Wentz F J, Krimchansky S and Skofronick-Jackson G M. 2015. The Global Precipitation Measurement (GPM) Microwave Imager (GMI): instrument overview and early on-orbit performance. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 8(7): 3452-3462
  10. 10.
    Fan C Q, Lin M Y, Zhao X G, Xie L Z, Wei L and Guo P. 2018. Research on parallelization of Fengyun satellite precipitation estimation daily end product algorithm//The 35th Annual Conference of the Chinese Meteorological Society S21 Satellite Meteorology and Ecological Remote Sensing. Hefei: [s.n.]: 33-38
  11. 11.
    Friedman J H. 2001. Greedy function approximation: a gradient boosting machine. The Annals of Statistics, 29(5): 1189-1232
  12. 12.
    Gao H, Tang S H and Han X Z. 2021. China’s Fengyun (FY) meteorological satellites, development and applications. Science and Technology Review, 39(15): 9-22
  13. 13.
    He Q, Li N, Luo W J and Shi Z Z. 2014. A survey of machine learning algorithms for big data. Pattern Recognition and Artificial Intelligence, 27(4): 327-336
  14. 14.
    Ke G L, Meng Q, Finley T, Wang T F, Chen W, Ma W D, Ye Q W and Liu T Y. 2017. LightGBM: a highly efficient gradient boosting decision tree//Proceedings of the 31st International Conference on Neural Information Processing Systems. Long Beach: Curran Associates Inc.: 2149-3157
  15. 15.
    Li X T, Li Z L and Han R C. 2023. Evaluations of different bias correction methods on the GCM precipitation data. Journal of China Hydrology, 43(3): 93-100, 117
  16. 16.
    Liao R W, Zhang D B and Shen Y. 2015. Validation of six satellite-derived rainfall estimates over China. Meteorological Monthly, 41(8): 970-979
  17. 17.
    Liu S X, Xia J and Wo X G. 2005. Advances in predictions in ungauged basins. Water Resources and Hydropower Engineering, 36(2): 9-12
  18. 18.
    Liu Y B, Fu Q N, Song P, Zhao X S and Dou C C. 2011. Satellite retrieval of precipitation: an overview. Advances in Earth Science, 26(11): 1162-1172
  19. 19.
    Liu Z Y. 2009. Research progress and prospect of flood forecasting technology in China. China Flood and Drought Management, 19(5): 13-16
  20. 20.
    Long K J, Gu J X, Shi C X, Pan Y and Huang X L. 2020. Quality Assessment of several merged precipitation products in a heavy rainfall process in Sichuan. Plateau and Mountain Meteorology Research, 40(2): 31-37
  21. 21.
    Lu F, Zhang X H, Chen B Y, Liu H, Wu R H, Han Q, Feng X H, Li Y and Zhang Z Q. 2017. FY-4 geostationary meteorological satellite imaging characteristics and its application prospects. Journal of Marine Meteorology, 37(2): 1-12
  22. 22.
    Lundberg S M, Erion G G and Lee S I. 2019. Consistent individualized feature attribution for tree ensembles. arXiv:1802.03888
  23. 23.
    Nguyen G, Dlugolinsky S, Bobák M, Tran V, García Á L, Heredia I, Malík P and Hluchý L. 2019. Machine learning and deep learning frameworks and libraries for large-scale data mining: a survey. Artificial Intelligence Review, 52(1): 77-124
  24. 24.
    Ren Y J, Yong B, Lu D K and Chen H Q. 2019. Evaluation of the Integrated Multi-satellitE Retrievals (IMERG) for Global Precipitation Measurement (GPM) mission over the Mainland China at multiple scales. Journal of Lake Sciences, 31(2): 560-572
  25. 25.
    Shinde P P and Shah S. 2018. A review of machine learning and deep learning applications//2018 Fourth International Conference on Computing Communication Control and Automation (ICCUBEA). Pune: IEEE: 1-6
  26. 26.
    Skofronick-Jackson G, Petersen W A, Berg W, Kidd C, Stocker E F, Kirschbaum D B, Kakar R, Braun S A, Huffman G J, Iguchi T, Kirstetter P E, Kummerow C, Meneghini R, Oki R, Olson W S, Takayabu Y N, Furukawa K and Wilheit T. 2017. The Global Precipitation Measurement (GPM) mission for science and society. Bulletin of the American Meteorological Society, 98(8): 1679-1695
  27. 27.
    Smith E A, Asrar G, Furuhama Y, Ginati A, Mugnai A, Nakamura K, Adler R F, Chou M D, Desbois M, Durning J F, Entin J K, Einaudi F, Ferraro R R, Guzzi R, Houser P R, Hwang P H, Iguchi T, Joe P, Kakar R, Kaye J A, Kojima M, Kummerow C, Kuo K S, Lettenmaier D P, Levizzani V, Lu N M, Mehta A V, Morales C, Morel P, Nakazawa T, Neeck S P, Okamoto K, Oki R, Raju G, Shepherd J M, Simpson J, Sohn B J, Stocker E F, Tao W K, Testud J, Tripoli G J, Wood E F, Yang S and Zhang W J. 2007. International Global Precipitation Measurement (GPM) program and mission: an overview//Levizzani V, Bauer P and Turk F J, eds. Measuring Precipitation from Space: Eurainsat and the Future. Dordrecht: Springer: 611-653
  28. 28.
    Song L. 2015. Study of Regional Precipitation Product with High Spatial-Temporal Resolution over the Tibetan Plateau based on TRMM 3B43. Nanjing: Nanjing University of Information Science and Technology
  29. 29.
    Tang G Q, Wan W, Zeng Z Y, Guo X L, Li N, Di L and Hong Y. 2015. An overview of the Global Precipitation Measurement (GPM) mission and it’s latest development. Remote Sensing Technology and Application, 30(4): 607-615
  30. 30.
    Tang S H and Mao L Y. 2020. “Fengyun” application for 50 years: leapfrog development, serve the world. Satellite Application, (11): 8-13
  31. 31.
    Tian H. 2021. Typhoon Rainfall Inversion Technology and Error Analysis based on Fengyun-4 Sensor. Nanjing: Nanjing University of Information Science and Technology
  32. 32.
    Wang C. 2019. Research on Quality Evaluation and Correction Method of Satellite Telemetry Precipitation Data. Chongqing: Chongqing Jiaotong University
  33. 33.
    Wang G Q. 2004. Fengyun-4 meteorological satellite multi-channel scanning imaging radiometer//Mr. Daheng’s 90th Birthday Collection and Proceedings of the 2004 Academic Conference of the Chinese Optical Society. Hangzhou: Zhejiang University Press: 1436-1439
  34. 34.
    Wei Y X. 2022. Downscaling Correction of Multi-Source Satellite Remote Sensing Precipitation Products in Pingtang Basin and its Applicability in Runoff Simulation. Nanning: Guangxi University
  35. 35.
    Xu J Y and Yang Y. 2018. A survey of ensemble learning approaches. Journal of Yunnan University (Natural Sciences Edition), 40(6): 1082-1092
  36. 36.
    Yu L and Wu T J. 2004. Assemble learning: a survey of Boosting algorithms. Pattern Recognition and Artificial Intelligence, 17(1): 52-59
  37. 37.
    Zeng S K and Yong B. 2019. Evaluation of the GPM-based IMERG and GSMaP precipitation estimates over the Sichuan region. Acta Geographica Sinica, 74(7): 1305-1318
  38. 38.
    Zhai P M and Pan X H. 2003. Change in extreme temperature and precipitation over Northern China during the second half of the 20th century. Acta Geographica Sinica, 58(S1): 1-10
  39. 39.
    Zhang H Y and Tang B H. 2021. Remote sensing retrieval of total precipitable water under clear-sky atmosphere from FY-4A AGRI data by combining physical mechanism and random forest algorithm. National Remote Sensing Bulletin, 25(8): 1836-1847
  40. 40.
    Zhang J Y. 2010. Review and reflection on China's hydrological forecasting techniques. Advances in Water Science, 21(4): 435-443
  41. 41.
    Zhang Q, Tu M H, Ma S Q, Yang Z B and Luo Y C. 2007. Quality assessment of the observational data of automatic precipitation stations in China. Journal of Applied Meteorological Science, 18(3): 365-372
  42. 42.
    Zhang R, Yong B and Zeng S K. 2021. Evaluation of GPM satellite precipitation products over Mainland China. Yangtze River, 52(5): 50-59
  43. 43.
    Zhong Y L. 2021a. Based on the Observation of the Fengyun-4 Satellite AGRI, a Random Forest Algorithm is Used to Invert Ground Precipitation. Nanjing: Nanjing University of Information Science and Technology
  44. 44.
    Zhong Y L. 2021b. Evaluation and verification of FY-4A satellite quantitative precipitation estimation product. Journal of Agricultural Catastropholgy, 11(3): 96-98
  45. 45.
    Zuo H C, Lü S H and Hu Y Q. 2004. Variations trend of yearly mean air temperature and precipitation in China in the last 50 years. Plateau Meteorology, 23(2): 238-244

Lire l'article complet

The above content is generated by Large Model Translation. The translated content is for reference only. We do not assume any commercial or legal responsibilty for any consequences arising from the use of our website