An improved model for short-term qualitative rainfall prediction combined with GNSS PWV and meteorological parameters

doi:10.11947/j.AGCS.2024.20230064.

Abstract

Abstract:

With the progress of GNSS data processing technology and the improvement of the accuracy of its derived water vapor products, rich information of water vapor contained in GNSS PWV (precipitable water vapor) has been gradually applied to precipitation forecast. Due to the limited ability of the current short-range rainfall forecasting model, which combines GNSS PWV and meteorological parameters in mining parameter information, this paper proposes an improved model based on RF algorithm and the use of the variations and anomalies of PWV, temperature, pressure and relative humidity as model inputs. Applying the new model to Hubei, Hunan and Jiangxi provinces, the model performance is assessed and compared with the BPNN algorithm. The results show that, compared with the BPNN algorithm, the rainfall forecast correct rate of the new method rises from 87.28% to 89.57% while the false rate is reduced from 17.82% to 15.06%. Thereby, the new model can better capture the influence of PWV and meteorological parameters on rainfall. During periods of frequent severe rains, the new model performs even better with an increase of correct rate by 5.57% and a decrease of false rate by 2.37%. Further studies reveal that setting the forecast starting time at the current epoch t and the previous epoch t-1 to make a forecast at time t+1, the correct rate is increased slightly, but the false rate is also increased marginally.

Key words: GNSS PWV, qualitative rainfall forecast, anomaly departure

CLC Number:

P258

Zhaohui XIONG, Dunyong ZHENG, Yibin YAO, Changyong HE, Sichun LONG, Shide LU, Jian ZHOU, Xiangen LAI. An improved model for short-term qualitative rainfall prediction combined with GNSS PWV and meteorological parameters[J]. Acta Geodaetica et Cartographica Sinica, 2024, 53(10): 1981-1992.

Figures/Tables 16

Fig.1

Fig.2

Fig.3

Fig.4

Fig.5

Tab.1

Tab.2

Fig.6

Tab.3

Tab.4

Fig.7

Fig.8

Tab.5

Tab.6

Fig.9

Fig.10

References 30

[1]	OIKONOMOU C, TYMVIOS F, PIKRIDAS C, et al. Tropospheric delay performance for GNSS integrated water vapor estimation by using GPT2w model, ECMWF's IFS operational model and in situ meteorological data[J]. Advances in Geosciences, 2018, 45:363-375.
[2]	KATSOUGIANNOPOULOS S, PIKRIDAS C, ZINAS N, et al. Analysis of precipitable water estimates using permanent GPS station data during the Athens heavy rainfall on February 22th 2013[M]//RIZOS C, WILLIS P, eds. International Association of Geodesy Symposia. Cham: Springer International Publishing, 2015: 407-414.
[3]	周康辉. 基于深度卷积神经网络的强对流天气预报方法研究[D]. 北京: 中国气象科学研究院, 2021.
	ZHOU Kanghui. Research on severe convective weather forecast method based on deep convolution neural network[D]. Beijing: Chinese Academy of Meteorological Sciences, 2021.
[4]	BENEVIDES P, CATALAO J, MIRANDA P M A. On the inclusion of GPS precipitable water vapour in the nowcasting of rainfall[J]. Natural Hazards and Earth System Sciences, 2015, 15(12):2605-2616.
[5]	RAHIMI Z, MOHD SHAFRI H Z, NORMAN M. A GNSS-based weather forecasting approach using nonlinear auto regressive approach with exogenous input (NARX)[J]. Journal of Atmospheric and Solar-Terrestrial Physics, 2018, 178:74-84.
[6]	LI Longjiang, ZHANG Kefei, WU Suqin, et al. An improved method for rainfall forecast based on GNSS-PWV[J]. Remote Sensing, 2022, 14(17):4280.
[7]	KANDA M. Correlation between convective heavy rainfalls and GPS precipitable water[C]//Procedings of 2003 International Workshop on GPS Meteorology. Tsukuba: [s.n.], 2003: 3-12.
[8]	刘旭春, 王艳秋, 张正禄. 利用GPS技术遥感哈尔滨地区大气可降水量的分析[J]. 测绘通报, 2006(4):10-12, 16.
	LIU Xuchun, WANG Yanqiu, ZHANG Zhenglu. Analysis of Harbin area atmosphere precipitable water vapor by using GPS technology[J]. Bulletin of Surveying and Mapping, 2006(4):10-12, 16.
[9]	王勇, 何荣, 杨彬云, 等. GPS反演的可降水量与降水的对比分析研究[J]. 测绘科学, 2010, 35(5):80-82.
	WANG Yong, HE Rong, YANG Binyun, et al. Study of comparisons between GPS precipitable water vapor and rainfall[J]. Science of Surveying and Mapping, 2010, 35(5):80-82.
[10]	SHI Junbo, XU Chaoqian, GUO Jiming, et al. Real-time GPS precise point positioning-based precipitable water vapor estimation for rainfall monitoring and forecasting[J]. IEEE Transactions on Geoscience and Remote Sensing, 2015, 53(6):3452-3459.
[11]	SUPARTA W, ZAINUDIN S K. Precipitation analysis using GPS meteorology over Antarctic Peninsula[C]//Proceedings of 2015 International Conference on Space Science and Communication. Langkawi: IEEE, 2015: 493-497.
[12]	WANG Binyan, ZHAO Linna, BAI Xuemei. The characteristics investigation of ground-based GPS/PWV during the “7.21” extreme rainfall event in Beijing[C]//Proceedings of China Satellite Navigation Conference (CSNC) 2015 Proceedings: Volume II. Berlin: Springer, 2015: 563-574.
[13]	YAO Yibin, SHAN Lulu, ZHAO Qingzhi. Establishing a method of short-term rainfall forecasting based on GNSS-derived PWV and its application[J]. Scientific Reports, 2017, 7(1):12465.
[14]	ZHAO Qingzhi, YAO Yibin, YAO Wanqiang. GPS-based PWV for precipitation forecasting and its application to a typhoon event[J]. Journal of Atmospheric and Solar-Terrestrial Physics, 2018, 167:124-133.
[15]	刘洋, 赵庆志, 姚顽强. 联合GNSS PWV与气象参数的短临降雨预测[C]//第十一届中国卫星导航年会. 成都: [s.n.], 2020: 6.
	LIU Yang, ZHAO Qingzhi, YAO Wanqiang. Combining GNSS-derived PWV and meteorological parameters for short-term rainfall forecasting[C]//Proceedings of the 11th China Satellite Navigation Annual Conference. Chengdu: [s.n.], 2020: 6.
[16]	KIM Y J, JEE J B, LIM B. Investigating the influence of water vapor on heavy rainfall events in the southern Korean peninsula[J]. Remote Sensing, 2023, 15(2):340.
[17]	ROSE M S, SUNIL P S, ZACHARIA J, et al. Early detection of heavy rainfall events associated with the monsoon in Kerala, India using GPS derived ZTD and PWV estimates: a case study[J]. Journal of Earth System Science, 2023, 132(1):23.
[18]	LIU Yang, ZHAO Qingzhi, YAO Wanqiang, et al. Short-term rainfall forecast model based on the improved BP-NN algorithm[J]. Scientific Reports, 2019, 9(1):19751.
[19]	LI Haobo, WANG Xiaoming, ZHANG Kefei, et al. A neural network-based approach for the detection of heavy precipitation using GNSS observations and surface meteorological data[J]. Journal of Atmospheric and Solar-Terrestrial Physics, 2021, 225:105763.
[20]	GORISHNIY Y, RUBACHEV I, KHRULKOV V, et al. Revisiting deep learning models for tabular data[EB/OL]. [2023-03-08]. http://arxiv.org/abs/2106.11959.
[21]	LIANG Hong, CAO Yunchang, WAN Xiaomin, et al. Meteorological applications of precipitable water vapor measurements retrieved by the national GNSS network of China[J]. Geodesy and Geodynamics, 2015, 6(2):135-142.
[22]	SAASTAMOINEN J. Atmospheric correction for the troposphere and stratosphere in radio ranging satellites[M]//Geophysical Monograph Series. Washington, D. C.: American Geophysical Union, 2013: 247-251.
[23]	SAASTAMOINEN J. Introduction to practical computation of astronomical refraction[J]. Bulletin Géodésique (1946-1975), 1972, 106(1):383-397.
[24]	XIONG Zhaohui, ZHANG Bao, SANG Jizhang, et al. Fusing precipitable water vapor data in China at different timescales using an artificial neural network[J]. Remote Sensing, 2021, 13(9):1720.
[25]	李浩博. 基于GNSS大气反演信息的短临极端天气预警预报研究[D]. 徐州: 中国矿业大学, 2021.
	LI Haobo. Research on early warning and forecasting of short-term and impending extreme weather based on GNSS atmospheric inversion information[D]. Xuzhou: China University of Mining and Technology, 2021.
[26]	ZHAO Qingzhi, LIU Yang, YAO Wanqiang, et al. Hourly rainfall forecast model using supervised learning algorithm[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60:4100509.
[27]	MANANDHAR S, DEV S, LEE Y H, et al. A data-driven approach to detect precipitation from meteorological sensor data[C]//Proceedings of 2018 IEEE International Geoscience and Remote Sensing Symposium. Valencia: IEEE, 2018: 3872-3875.
[28]	YEN S J, LEE Yueshi. Cluster-based under-sampling approaches for imbalanced data distributions[J]. Expert Systems with Applications, 2009, 36(3):5718-5727.
[29]	RAHMANM M, DAVIS D. Cluster based under-sampling for unbalanced cardiovascular data[J]. Proceedings of the World Congress on Engineering, 2013, 3, 3-5.
[30]	MYERS-BEAGHTON A K, VVEDENSKY D D. Chapman-Kolmogorov equation for Markov models of epitaxial growth[J]. Journal of Physics A: Mathematical and General, 1989, 22(11):L467-L475.

模型	决策树数量	模型	决策树数量
RF_A_t	75	RF_B_t-1	80
RF_B_t	75	RF_C_t-1	95
RF_C_t	95	RF_D_t-1	85
RF_D_t	85

模型	Ts	正确率/(%)	错报率/(%)	漏报率/(%)
BPNN_A_t	0.73	87.3	17.8	12.7
BPNN_B_t	0.75	88.3	16.5	11.7
BPNN_C_t	0.75	88.4	16.3	11.6
BPNN_D_t	0.76	88.5	16.1	11.5
BPNN_B_t-1	0.75	88.4	16.8	11.6
BPNN_C_t-1	0.75	88.5	16.6	11.5
BPNN_D_t-1	0.76	88.7	16.3	11.3
RF_A_t	0.77	89.0	15.5	11.1
RF_B_t	0.77	89.3	15.4	10.7
RF_C_t	0.77	89.3	15.2	10.7
RF_D_t	0.77	89.6	15.1	10.4
RF_B_t-1	0.77	89.6	16.0	10.4
RF_C_t-1	0.77	89.6	15.8	10.4
RF_D_t-1	0.77	89.8	15.7	10.2

模型	Ts变化	正确率变化/(%)	漏报率变化/(%)	错报率变化/(%)
RF_A_t	0.03	1.7	-2.3	-1.7
RF_B_t	0.03	2.0	-2.4	-2.0
RF_C_t	0.04	2.0	-2.6	-2.0
RF_D_t	0.04	2.3	-2.8	-2.3
RF_B_t-1	0.03	2.3	-1.8	-2.3
RF_C_t-1	0.03	2.3	-2.0	-2.3
RF_D_t-1	0.04	2.5	-2.1	-2.5

模型	Ts评分	正确率/(%)	错报率/(%)	漏报率/(%)
BPNN_A_t	0.66	78.0	18.4	22.0
RF_A_t	0.70	81.8	17.1	18.2
RF_D_t	0.72	83.6	15.6	16.5

模型	Ts评分	正确率/(%)	错报率/(%)	漏报率/(%)
BPNN_A_t	0.70	83.4	18.5	16.7
RF_A_t	0.73	84.7	16.6	15.3
RF_D_t	0.74	85.3	15.5	14.7