长短期记忆神经网络(LSTM)模型在低能见度预报中的应用

doi:10.3969/j.issn.1673-503X.2022.05.004

Abstract

Abstract:

Based on the hourly meteorological observation data (relative humidity, wind speed, ground-air temperature difference, and visibility) and air quality index (AQI) data from 2015 to 2019 in Yiwu of Zhejiang province, we analyzed the distribution and meteorological conditions of low visibility (observed visibility lt; 10 km) in Yiwu.The Long Short-term Memory Neural Network (LSTM) was used to simulate the hourly visibility, and the simulation results with and without observed visibility as an input member were compared.According to the meteorological conditions of low visibility, the simulation period was divided into three periods (from November to February, from March to June, and from July to October).We compared the simulation performance in different periods and evaluated the model prediction steps.The results showed that the high humidity, high pollution, air temperature higher than the surface temperature, and low wind speed are the main meteorological characteristics influencing the low visibility weather in Yiwu.The LSTM has a good performance in simulating visibility at a single station.When the historical observed visibility has been used as an input parameter, the simulation accuracy can be greatly improved, with the root mean square error (RMSE) of 0.63 km, the mean absolute error (MAE) of 0.51 km, and the fitting goodness R² of 0.99.Better simulation can be reached after dividing different periods, and the best simulation occurred in winter (from November to February) using the input elements selected in this study, with RMSE=2.35 km and MAE=1.46 km, and for low visibility weather, the RMSE_10km=1.81 km, MAE_10km=1.13 km, and R²=0.83.In the simulation from March to June, the simulation of low visibility without AQI as one input member performs better, which means that low visibility in Yiwu is dominated by foggy weather in this period.Adding too many variables does not necessarily improve the accuracy of the model.With the increase in the forecast step size, the simulation performs worse.When the forecast step size equals 3 h, R² reaches 0.71 and the simulation result has no practical application significance.

Key words: Atmospheric Visibility, Long Short-term Memory Neural Network (LSTM), Forecast model, Air quality

CLC Number:

P427.2

Nan FANG, Guo-quan XIE, Xiao-jian RUAN, Chen-ping REN, Shu-jie JIANG, Wei-wei ZHANG. Application of long short-term memory neural network (LSTM) model in low visibility forecast[J]. Journal of Meteorology and Environment, 2022, 38(5): 34-41.

Figures/Tables 9

Fig.1

Table 1

Fig.2

Fig.3

Fig.4

Fig.5

Fig.6

Fig.7

Fig.8

References 28

1	傅刚, 李晓岚, 魏娜. 大气能见度研究[J]. 中国海洋大学学报, 2009, 39 (5): 855- 862.
2	李生才, 笑蕾. 2019年1—2月国内生产安全事故统计分析[J]. 安全与环境学报, 2019, 19 (2): 722- 724. doi: 10.13637/j.issn.1009-6094.2019.02.050
3	白志鹏, 蔡斌彬, 董海燕, 等. 灰霾的健康效应[J]. 环境污染与防治, 2006, 28 (3): 198- 201. doi: 10.3969/j.issn.1001-3865.2006.03.012
4	倪江波, 李文才, 尚可政, 等. 华北区域性低能见度天气的自动识别及预报[J]. 干旱气象, 2015, 33 (1): 174- 179.
5	魏春璇, 黄鹤, 翟振芳, 等. 合肥市霾日气象要素特征分析及预报方法研究[J]. 大气与环境光学学报, 2019, 14 (6): 419- 430.
6	夏凡, 李昌义. 基于3种能见度方案山东地区雾天气预报试验研究[J]. 气象与环境学报, 2018, 34 (3): 48- 57. doi: 10.3969/j.issn.1673-503X.2018.03.006
7	王恺, 赵宏, 刘爱霞, 等. 基于风险神经网络的大气能见度预测[J]. 中国环境科学, 2009, 29 (10): 1029- 1033. doi: 10.3321/j.issn:1000-6923.2009.10.005
8	王勇. 基于多源数据和XGBoost算法的上海市能见度预测模型研究[D]. 上海: 华东师范大学, 2019.
9	Kurt A , Oktay A B . Forecasting air pollutant indicator levels with geographic models 3 days in advance using neural networks[J]. Expert Systems with Applications, 2010, 37 (12): 7986- 7992. doi: 10.1016/j.eswa.2010.05.093
10	孙全德, 焦瑞莉, 夏江江, 等. 基于机器学习的数值天气预报风速订正研究[J]. 气象, 2019, 45 (3): 426- 436.
11	毛亚萍, 房世峰. 基于机器学习的参考作物蒸散量估算研究[J]. 地球信息科学学报, 2020, 22 (8): 1692- 1701.
12	Graves A , Schmidhuber J . Framewise phoneme classification with bidirectional LSTM and other neural network architectures[J]. Neural Networks, 2005, 18 (5/6): 602- 610.
13	杨丽, 吴雨茜, 王俊丽, 等. 循环神经网络研究综述[J]. 计算机应用, 2018, 38 (S2): 1- 6.1-6, 26
14	梁军, 柴玉梅, 原慧斌, 等. 基于极性转移和LSTM递归网络的情感分析[J]. 中文信息学报, 2015, 29 (5): 152- 159.
15	陶晔, 杜景林. 基于随机森林的长短期记忆网络气温预测[J]. 计算机工程与设计, 2019, 40 (3): 737- 743.
16	王炜, 刘宏伟, 陈永杰, 等. 基于LSTM循环神经网络的风力发电预测[J]. 可再生能源, 2020, 38 (9): 1187- 1191.
17	朱乔木, 李弘毅, 王子琪, 等. 基于长短期记忆网络的风电场发电功率超短期预测[J]. 电网技术, 2017, 41 (12): 3797- 3802.
18	刘新, 赵宁, 郭金运, 等. 基于LSTM神经网络的青藏高原月降水量预测[J]. 地球信息科学学报, 2020, 22 (8): 1617- 1629.
19	刘蕾. 基于Tensorflow的循环神经网络模型在上海市空气质量预测中的应用[D]. 上海: 上海师范大学, 2019.
20	傅昭娟. "十二五"期间义乌市环境空气质量及影响因素分析研究[J]. 环境科学与管理, 2016, 41 (8): 116- 117.116-117, 134
21	俞江仙. 义乌市大气污染状况调查[J]. 环境污染与防治, 2000, 22 (2): 28- 30.
22	李佳润. 义乌市乡村经济特色及发展质量研究[D]. 舟山: 浙江海洋大学, 2020.
23	董佳丹, 陈晓玲, 蔡晓斌, 等. 基于中国大气环境监测站点的2015—2019年大气质量状况时空变化分析[J]. 地球信息科学学报, 2020, 22 (10): 1983- 1995.
24	环境保护部. 环境空气质量指数(AQI)技术规定: HJ 633-2012[S]. 北京: 中国环境科学出版社, 2016.
25	Hochreiter S , Schmidhuber J . Long short-term memory[J]. Neural Computation, 1997, 9 (8): 1735- 1780.
26	Kingma D P, Ba L J. Adam: A method for stochastic optimization[C]//3rd International Conference on Learning Representations. San Diego: ICLR, 2015: 116-130.
27	王博妮, 张雪蓉, 孙明, 等. 江苏地区雨雾天气特征及成因研究[J]. 气象与环境学报, 2020, 36 (1): 58- 66.
28	俞燎霓, 胡波, 曹美兰, 等. 浙江梅雨气候特征及其梅汛期暴雨的形势分析[J]. 科技通报, 2010, 26 (5): 782- 786.782-786, 794

空气质量分指数(IAQI)	二氧化硫(SO₂)		二氧化氮(NO₂)		一氧化碳(CO)		臭氧(O₃)		颗粒物(粒径小于等于2.5 μm)
空气质量分指数(IAQI)	24 h平均/(μg·m^-3)	1 h平均/(μg·m^-3)^①	24 h平均/(μg·m^-3)	1 h平均/(μg·m^-3)^①	24 h平均/(mg·m^-3)	1 h平均/(mg·m^-3)^①	1 h平均/(μg·m^-3)	8 h滑动平均/(μg·m^-3)	24 h平均/(μg·m^-3)	24 h平均/(μg·m^-3)
0	0	0	0	0	0	0	0	0	0	0
50	50	150	40	100	2	5	160	100	50	35
100	150	500	80	200	4	10	200	160	150	75
150	475	650	180	700	14	35	300	215	250	115
200	800	800	280	1200	24	60	400	265	350	150
300	1600	②	565	2340	36	90	800	800	420	250
400	2100	②	750	3090	48	120	1000	③	500	350
500	2620	②	940	3840	60	150	1200	③	600	500

[1]	Xiao-hui MA,Jing XU,Zi-yin ZHANG,Zi-ming LI,Wei WEN. Influence of the ultra-high voltage cross-regional transmission on air quality [J]. Journal of Meteorology and Environment, 2022, 38(4): 136-144.
[2]	Wei GUO, Ling-yun ZHU, Wen-ya WANG, Xing-ai GAO, Peng-wei CHENG, Yue-jun ZHANG. Evaluation of air quality improvement in Taiyuan during Second National Youth Games [J]. Journal of Meteorology and Environment, 2021, 37(6): 36-43.
[3]	Min XU. Numerical simulation on the spatiotemporal distribution of PM_2.5 in the Yangtze River Delta [J]. Journal of Meteorology and Environment, 2021, 37(4): 33-39.
[4]	Xiao-lan LI, Wei-jun QUAN, Dong-dong WANG, Di WANG. Comparison of visibility measurements using DNQ1 and FD12 visibility meters and their major influencing factors [J]. Journal of Meteorology and Environment, 2021, 37(3): 25-32.
[5]	Yu ZHOU,Hong-xing HAN,Chen-xiao SHI,Jun XING,Li SUN. Comparative analysis of aerosol mass concentration characteristics in winter and spring during 2016 to 2017 in three cities of Hainan province [J]. Journal of Meteorology and Environment, 2020, 36(6): 82-90.
[6]	Yang-feng WANG,Yan-jun MA,Wei-jun QUAN,Rong-ping LI. Progress of the application research of air quality models in Northeast China [J]. Journal of Meteorology and Environment, 2020, 36(6): 130-136.
[7]	Hong-rui ZHU,He-nan LIU,Hong-ling ZHANG,Chang-jiao YIN. Change characteristics of atmospheric self-purification capability in Heilongjiang province from 2008 to 2018 [J]. Journal of Meteorology and Environment, 2020, 36(4): 89-94.
[8]	Xing-xing GAO, Hai-lin GUI, Liu-jie PAN, Nan WANG, Jian-peng WANG. Statistical analysis of air pollution in Fenwei Plain based on satellite remote sensing and ground observation data [J]. Journal of Meteorology and Environment, 2020, 36(4): 36-44.
[9]	Feng-xiang LONG, Yu-lin ZHANG. Relationship between atmospheric visibility and particulate matter concentration and meteorological parameters in Guilin urban area [J]. Journal of Meteorology and Environment, 2020, 36(1): 21-27.
[10]	ZHENG Qing-feng, SHI Jun, DONG Guang-tao, HUANG Wen-juan, XU Wei-zhong, WANG Quan-yuan. Characteristics of air quality index and its correlation with meteorological parameters in Shanghai [J]. Journal of Meteorology and Environment, 2019, 35(5): 53-62.
[11]	ZHAO Lei, WU Kun-ti, CHEN Ming. Characteristics of air quality in Haikou from 2013 to 2016 and pollutant source analysis during typical pollution events [J]. Journal of Meteorology and Environment, 2019, 35(5): 63-69.
[12]	YANG Dong, ZHU Jia-min, YAO Ri-sheng, TU Xiao-ping. Comparison of air quality and atmospheric self-purification capacity between coastal and inland areas of Ningbo [J]. Journal of Meteorology and Environment, 2019, 35(3): 52-59.
[13]	ZHANG Ying-ying, CHEN Ming, CHEN Li-ying. Potential source analysis of heavy pollution weather in Haikou [J]. Journal of Meteorology and Environment, 2019, 35(2): 107-111.
[14]	ZHU Hong-rui, LIU He-nan, ZHANG Hong-ling, YIN Chang-jiao. Characteristics of air quality and its relationship with meteorological factors in Harbin [J]. Journal of Meteorology and Environment, 2019, 35(1): 53-58.
[15]	DONG Zhen-hua, QI Yi-ling, KONG Hai-jiang. Comparison analysis of three heavy pollution events in He'nan province [J]. Journal of Meteorology and Environment, 2019, 35(1): 18-25.

Application of long short-term memory neural network (LSTM) model in low visibility forecast

RichHTML

PDF (PC)

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 9

References 28

Related Articles 15

Metrics

Comments

Recommended 10