Application of Machine Learning to Statistical Evaluation of Artificial Rainfall Enhancement

Li Dan; Lin Wen; Liu Qun; Feng Hongfang; Hu Shuping; Wang Zhihai

doi:10.11898/1001-7313.20240110

Vol.35, NO.1, 2024

Turn off MathJax

Article Contents

Article Navigation > Journal of Applied Meteorological Science > 2024 > 35(1): 118-128

Li Dan, Lin Wen, Liu Qun, et al. Application of machine learning to statistical evaluation of artificial rainfall enhancement. J Appl Meteor Sci, 2024, 35(1): 118-128. DOI: 10.11898/1001-7313.20240110.

Citation:

Li Dan, Lin Wen, Liu Qun, et al. Application of machine learning to statistical evaluation of artificial rainfall enhancement. J Appl Meteor Sci, 2024, 35(1): 118-128. DOI: 10.11898/1001-7313.20240110.

Citation:

PDF( 5574 KB)

Application of Machine Learning to Statistical Evaluation of Artificial Rainfall Enhancement

DOI: 10.11898/1001-7313.20240110

Li Dan^{1, 2, 3},
Lin Wen^{1, 2, 3
,
,},
Liu Qun⁴,
Feng Hongfang^{1, 2, 3},
Hu Shuping⁵,
Wang Zhihai⁴

1.
Fujian Institute of Meteorological Sciences, Fuzhou 350001
2.
Fujian Key Laboratory of Severe Weather, Fuzhou 350001
3.
Key Laboratory for Strait Disaster Weather of CMA, Fuzhou 350001
4.
School of Mathematics and Statistics, Minnan Normal University, Zhangzhou 363000
5.
Gutian Meteorological Station of Ningde, Fujian, Ningde 352000

Received Date: 2023-10-10
Rev Recd Date: 2023-12-01
Publish Date: 2024-01-31

Abstract

Abstract

As an important part of weather modification operation, the scientific effectiveness evaluation of artificial rainfall enhancement has gradually attracted attentions of government and public. In order to evaluate effects of artificial rainfall enhancement objectively and quantitatively, combing linear fitting, polynomial regression, spline regression and 3 other machine learning methods including decision tree, support vector machine and neural network, the relationship model between the rainfall in the target area and the contrast area is established based on rainfall data and operation information of recent 10 years in Fujian. Different historical regression statistical test schemes of rainfall enhancement effects are compared and analyzed, aiming to further optimize the best natural rainfall estimation algorithm based on alterable contrast area with statistical method, which can provide reference for the assessment of artificial rainfall enhancement effects. Results show that historical rainfall data samples are mainly concentrated in the weak rainfall grade. Using multiple regression methods (linear regression, polynomial regression and spline regression), the piecewise statistics of rainfall data does not significantly improve the linear regression model between two regions, and its root mean square error (RMSE) is generally higher than the statistical results. By comparing various machine learning and linear regression models, it is found that CNN and quomial regression perform relatively well when the regional average surface rainfall is taken as the statistical variable, with the determination coefficient of CNN being 0.516 and RMSE being 1.097 mm. Each statistical model is greatly improved after six root square transformations of rainfall data. The performance of the model established by CNN method is relatively optimal, with determination coefficient R² up to 0.658 and RMSE only , followed by SVM statistical model based on sixth-order root square transform data, with R² being 0.41 and RMSE being . In order to further overcome the time asynchronization and uneven spatial distribution of rainfall in the two regions, the convolutional neural network CNN optimizers (RMSP, ADAM and SGD) are used to establish the contrast-target region rainfall relationship model based on the grid data of natural rainfall plane. Comparison results show that the ADAM optimizer model is the best with the RMSE of 0.61 mm, and its ability to estimate natural rainfall in the affected area is enhanced, by which method the disturbance of the heavy rainfall center will be reduced.
- artificial rainfall effect evaluation;
- regional historical regression;
- machine learning;
- statistical test

FullText(HTML)

References(42)

Figures and Tables

Fig. 1 Gutian artificial rainfall effect evaluation base in Fujian

DownLoad: Full-Size Img PowerPoint

Fig. 2 Fitting relationship between topography similarity and rainfall correlation coefficient in target area and contrast area

DownLoad: Full-Size Img PowerPoint

Fig. 3 Artificial rainfall target area, floating contrast area and the best-contrast area of Gutian

DownLoad: Full-Size Img PowerPoint

Fig. 4 Sample size of different rainfall intensity

DownLoad: Full-Size Img PowerPoint

Fig. 5 Root mean square error of rainfall estimation for different rainfall intensity by different models

DownLoad: Full-Size Img PowerPoint

Fig. 6 Root mean square error and determination coefficient of rainfall estimation based on total samples by different models

DownLoad: Full-Size Img PowerPoint

Fig. 7 Regional grid and automatic weather station in target area and contrast area

DownLoad: Full-Size Img PowerPoint

Fig. 8 Area rainfall distribution with different methods at 1900 BT 9 Apr 2016

DownLoad: Full-Size Img PowerPoint

Fig. 9 Comparison of rainfall estimation among 3 machine learning optimizers

DownLoad: Full-Size Img PowerPoint

Table 1 Sample size of different rainfall categories in 2014-2023

降水等级	雨强/(mm·h^-1)	样本量	样本比例/%
弱降水	[0.1, 5)	25207	95.98
一般降水	[5, 10)	780	2.97
中等降水	[10, 25)	269	1.02
强降水	[25, +∞)	7	0.03

DownLoad: Download CSV

References

[1]	Mao J T, Zheng G G. Discussions on some weather modification issues. J Appl Meteor Sci, 2006, 17(5): 643-646. doi: 10.3969/j.issn.1001-7313.2006.05.015
[2]	Guo X L, Fang C G, Lu G X, et al. Progresses of weather modification technologies and applications in China from 2008 to 2018. J Appl Meteor Sci, 2019, 30(6): 641-650. doi: 10.11898/1001-7313.20190601
[3]	Yao Z Y. Review of weather modification research in Chinese Academy of Meteorological Sciences. J Appl Meteor Sci, 2006, 17(6): 786-795. doi: 10.3969/j.issn.1001-7313.2006.06.016
[4]	Ye J D, Fan B F. Statistical and Mathematical Methods of Weather Modification. Beijing: Science Press, 1982.
[5]	Ye J D, Fan B F, Du J C. Study of negative effects in artificial precipitation enhancement experiments. J Appl Meteor Sci, 1998, 9(3): 336-344. http://qikan.camscma.cn/article/id/19980348
[6]	Zeng G P, Fang S Z. The result multivariate analysis of artificial rainfall in Fujian Gutian Area during 1975-1984. J Trop Meteor, 1986, 2(4): 336-342.
[7]	Zeng G P, Fang S Z, Xiao F. The total analysis of the effect of artificial rainfall in Gutian Reservoir Area, Fujian(1975-1986). Chinese J Atmos Sci, 1991, 15(4): 97-108. doi: 10.3878/j.issn.1006-9895.1991.04.11
[8]	Zeng G P, Wu M L, Lin C C, et al. A comprehensive evaluation of the effect of artificial precipitation in Gutian Reservior Area. J Appl Meteor Sci, 1993, 4(2): 154-161. http://qikan.camscma.cn/article/id/19930229
[9]	Zeng G P. Statistical simulation study on the effect of non-randomized artificial precipitation enhancement experiment. J Appl Meteor Sci, 1999, 10(2): 255-256. doi: 10.3969/j.issn.1001-7313.1999.02.017
[10]	Zeng G P, Zhang C A, Li M L. Study on statistic numerical simulation method of precept statistic design of artificial precipitation. Chinese J Atmos Sci, 2000, 24(1): 131-141.
[11]	Jiang N C, Wu L L, Zeng G P. On effect test of drought-resistant rocketry artificial precipitation enhancement operation. Meteor Mon, 2006, 32(8): 54-58.
[12]	Wang Y L, Li D S, Liu S J. Stratified sampling historical regression method for aircraft precipitation enhancement effect test. Climate Environ Res, 2012, 17(6): 862-870.
[13]	Wang W, Shi Y H, Li H Y, et al. A method for evaluating effectiveness of convective cloud precipitation enhancement and its application. Meteor Sci Technol, 2014, 42(6): 1131-1136. doi: 10.3969/j.issn.1671-6345.2014.06.031
[14]	Jia S, Yao Z Y. Case study on the convective clouds seeding effects in Yangtze-Huaihe Region. Meteor Mon, 2016, 42(2): 238-245.
[15]	Tang R M, Yuan Z T, Xiang Y C, et al. A method for selecting contrast cloud automatically based on radar echo in effectiveness evaluation of rain enhancement. Meteor Mon, 2010, 36(4): 96-100. doi: 10.3969/j.issn.1673-8411.2010.04.028
[16]	Wang Y L, Yao Z Y, Lin C C. Analysis of radar echoes at different heights before and after precipitation enhancement. J Arid Meteor, 2018, 36(4): 644-651.
[17]	Liu Q, Yao Z Y. On physical eveluation of aircraft cloud seeding and case study. Meteor Mon, 2013, 39(10): 1359-1368. doi: 10.7519/j.issn.1000-0526.2013.10.015
[18]	Hu S P, Lin W, Lin C C, et al. Physical inspection of randomized trial for the artificial rain enhancement experiment at Gutian from 2014 to 2022. J Appl Meteor Sci, 2023, 34(6): 706-716. doi: 10.11898/1001-7313.20230606
[19]	Lou X F, Fu Y, Sun J. A numerical seeding simulation of convective precipitation in Zhejiang, China. J Appl Meteor Sci, 2019, 30(6): 665-676. doi: 10.11898/1001-7313.20190603
[20]	Hong Y C, Zhou F F. A numerical simulation study of precipitation formation mechanism of "seeding-feeding" cloud system. Chinese J Atmos Sci, 2005, 29(6): 885-896. doi: 10.3878/j.issn.1006-9895.2005.06.05
[21]	Gong D L, Wang J, Liu S J. Numerical simulation of cloud microphysical structure and artificial seeding condition in precipitation cloud in Shandong Province. Plateau Meteor, 2006, 25(4): 723-730. doi: 10.3321/j.issn:1000-0534.2006.04.022
[22]	Wang W, Yao Z Y. Statistical estimation of artificial precipitation enhancement effectiveness in Beijing in 2006. Plateau Meteor, 2009, 28(1): 195-202.
[23]	Wang L, Wei Z A, Cheng P, et al. Statistical tests and analysis of effective evaluation of artificial precipitation enhancement operation of Hunan. J Meteor Res Appl, 2019, 40(3): 85-89. doi: 10.3969/j.issn.1673-8411.2019.03.020
[24]	Liu Q. The Statistical Method Optimization and Case Study of Effectiveness Test in Precipitation Enhancement. Beijing: Academy of Meteorological Sciences, 2013.
[25]	Cheng P, Chen Q, Jiang Y Y, et al. Effect evaluation of artificial rainfall enhancement in the Shiyang River Basin of Hexi Corridor in the latest 10 years. Plateau Meteor, 2021, 40(4): 866-874.
[26]	Wang F, Li J M, Yao Z Y, et al. Advances of quantitative evaluation studies of artificial precipitation enhancement in China. Meteor Mon, 2022, 48(8): 945-962.
[27]	Wang W, Yao Z Y. Accuracy analysis of statistical evaluation result in precipitation enhancement experiment. Meteor Sci Technol, 2009, 37(2): 209-215. doi: 10.3969/j.issn.1671-6345.2009.02.018
[28]	Ye J D, Li T L. Evaluation methods of cloud seeding effect with regional control and covariable regression analysis. Sci Meteor Sinica, 2001, 21(1): 64-72.
[29]	Wu X H, Niu S J, Jin D Z, et al. Influence of natural rainfall variability on the evaluation of artificial precipitation enhancement. Sci China(Earth Sci), 2015, 45(7): 1011-1019.
[30]	Fang B, Xiao H, Ban X X. Comparison between CA-FCM and some other methods for evaluating precipitation enhancement effectiveness. Meteor Sci Technol, 2008, 36(5): 612-621. doi: 10.3969/j.issn.1671-6345.2008.05.021
[31]	Fang B, Xiao H, Wang Z H, et al. Application of cluster analysis to the statistical assessment of the effect of artifical rain enhancement. J Nanjing Inst Meteor, 2005, 28(6): 739-745. doi: 10.3969/j.issn.1674-7097.2005.06.003
[32]	Zhai Y, Xiao H, Du B Y, et al. Application of the cluster statistical test to effectiveness evaluation of artificial precipitation enhancement. J Nanjing Inst Meteor, 2008, 31(2): 228-233. doi: 10.3969/j.issn.1674-7097.2008.02.012
[33]	Hu Y Y, Pang L, Wang Q G. Application of deep learning bias correction method to temperature grid forecast of 7-15 days. J Appl Meteor Sci, 2023, 34(4): 426-437. doi: 10.11898/1001-7313.20230404
[34]	Mi Q C, Gao X N, Li Y, et al. Application of deep learning method to drought prediction. J Appl Meteor Sci, 2022, 33(1): 104-114. doi: 10.11898/1001-7313.20220109
[35]	Liu H Z, Xu H, Bao H J, et al. Application of machine learning classification algorithm to precipitation-induced landslides forecasting. J Appl Meteor Sci, 2022, 33(3): 282-292. doi: 10.11898/1001-7313.20220303
[36]	Lan Y, Luo C, Wu Z F, et al. The assessment of application effectiveness of three machine learning methods in automatic identification of thunderstorm gale in Guangdong. J Trop Meteor, 2023, 39(2): 256-266.
[37]	Yin X Y, Hu Z Q, Zheng J F, et al. Filling in the dual polarization radar echo occlusion based on deep learning. J Appl Meteor Sci, 2022, 33(5): 581-593. doi: 10.11898/1001-7313.20220506
[38]	Wang X. Big Data Analysis: Methods and Applications. Beijing: Tsinghua University Press, 2013.
[39]	Hu Z J. Covariance statistical analysis method for testing the effect of artificial precipitation. Meteor Mon, 1979, 5(9): 31-33.
[40]	Lin C C, Yao Z Y, Lin W, et al. Analysis on cloud echoes characteristics and operational conditions of precipitation enhancement in Gutian of Fujian. Trans Atmos Sci, 2017, 40(1): 138-144.
[41]	Wang X, Chu T J. Non-parametric Statistics(2nd ed). Beijing: Tsinghua University Press, 2014.
[42]	Wang H J. Machine Learning: Python Sklearn Tensor Flow 2.0 Micro-Lesson Video Version. Beijing: Tsinghua University Press, 2020.