[1]郜庆林,简 单.基于卷积神经网络算法的金华地区高山气象观测站逐时气温预报订正[J].山地学报,2024,(2):278-286.[doi:10.16089/j.cnki.1008-2786.000822]
　GAO Qinglin,JIAN Dan.Hourly Temperature Forecast Revision Based on Convolutional Neural Network Algorithm on Observations at Alpine Meteorological Stations in Jinhua Area of China[J].Mountain Research,2024,(2):278-286.[doi:10.16089/j.cnki.1008-2786.000822]
点击复制

基于卷积神经网络算法的金华地区高山气象观测站逐时气温预报订正
分享到：

《山地学报》[ISSN:1008-2186/CN:51-1516]

卷:
期数:
2024年第2期
页码:
278-286
栏目:
山地技术
出版日期:
2024-03-25

文章信息/Info

Title:
Hourly Temperature Forecast Revision Based on Convolutional Neural Network Algorithm on Observations at Alpine Meteorological Stations in Jinhua Area of China
文章编号:
1008-2786-(2024)2-278-9
作者:
郜庆林1; 简 单2
(1.金华市气象局,浙江 金华 321000; 2.兰溪市气象局,浙江 兰溪 321100)
Author(s):
GAO Qinglin1; JIAN Dan2
(1. Jinhua Meteorological Bureau, Jinhua 321000, Zhejiang, China; 2. Lanxi Meteorological Bureau, Lanxi 321100, Zhejiang, China)
关键词:
卷积神经网络;  高山气象观测站;  精细气温;  订正预报;  智能网格;  金华
Keywords:
convolutional neural network;  alpine meteorological observatory;  fine temperature;  forecast revision;  intelligent grid;  Jinhua
分类号:
P423
DOI:
10.16089/j.cnki.1008-2786.000822
文献标志码:
A
摘要:
受地貌和气候条件影响,智能网格气温预报产品的预报结果,在复杂地形山区易出现误差。通过人工订正关键点降低误差,存在较大主观性,预报精度有限,难以满足精细化气象服务需求。神经网络算法能够大幅提升预报精度,然而这种方法在气温逐时波动小的平原地区应用较多,在气温波动大的山区鲜有应用。本文以浙江金华山区为研究区域,基于浙江省智能网格温度预报产品以及同时段高山气象观测站逐时气温观测数据,采用卷积神经网络算法,实现关键点气温逐时预报订正。研究结果表明:(1)小时尺度上,订正后各站点的气温均方根误差均显著减小,由订正前3 ℃～7 ℃减小至订正后2 ℃～3 ℃,订正后的预测结果更加准确,订正效果符合预期。(2)月尺度上,相较智能网格原始气温预报数据,该模型预报结果准确率也明显提升,订正后月平均气温准确率提高了33.18%～46.86%,其中准确率6月最高。(3)相较人工订正的方式,该模型对山地气温预报的订正能力更稳定,模型的两项关键指标(平均绝对误差和2 ℃预报准确率)均接近或超过同时段浙江省天气预报质量检验平台气温业务指标。该研究结果满足金华地区高山气象观测站对于气温预报产品的业务可用性需求,可为精细化山区气象服务提供数据支撑。
Abstract:
Smart grid temperature forecast products are subjected to geomorphology and climatic conditions, and the temperature forecast results are prone to errors in mountainous areas with complex terrain. Reducing the errors by manually revising observations at some key points of an alpine meteorological observatory network had a large subjectivity and limited forecast accuracy, making it difficult to meet the demand for refined meteorological services.
Neural network algorithms offer an objective computational approach to greatly improving forecast accuracy by learning from historical temperature data to regulate current data; however, this method was more commonly applied in plains with small hourly temperature fluctuations, and seldom used in mountainous areas where temperature fluctuations are large.
In this paper, the Jinhua mountainous area of Zhejiang province, China was taken as research target. The intelligent grid temperature forecast product of Zhejiang province were calibrated with simultaneous hourly temperature observations collected at alpine meteorological observation stations by convolutional neural network(CNN)algorithm for proper revisions on hourly temperature prediction at key points/stations.
(1)On an hourly scale, the root mean square error of temperatures at each station significantly decreased after the CNN revision, from 3 ℃-7 ℃ before the revision to 2 ℃-3 ℃ after the revision, satisfying expectation of accuracy.
(2)On a monthly scale, temperature prediction accuracy improved notably as compared with product by the intelligent grid. After the revision, the accuracy of the monthly average temperature increased by 33.18% to 46.86%, with the highest accuracy in June.
(3)The CNN model was more stable in revising mountain temperature forecasts than the manual revision method. It justified that two key indicators(average absolute error and 2 ℃ forecast accuracy)were close to or exceeded the concurrent acceptance check standards of Zhejiang Provincial Weather Forecast Quality Inspection Platform.
This study improves the operational availability of temperature forecast products in alpine meteorological observatory in Jinhua area, which can provide data support for refined mountain weather services in mountainous areas.

参考文献/References:

[1] 易桂花, 张廷斌, 何奕萱, 等. 四种气温空间插值方法适用性分析[J]. 成都理工大学学报(自然科学版), 2020, 47(1): 115-128. [YI Guihua, ZHANG Tingbin, HE Yixuan, et al. Applicability analysis of four spatial interpolation methods for air temperature [J]. Journal of Chengdu University of Technology(Science and Technology Edition), 2020, 47(1): 115-128] DOI: 10.3969/j.issn.1671-9727.2020.01.11
[2] 王莹, 苏永秀, 李政. 广西西部山区日最低气温短序列订正方法[J]. 山地学报, 2012, 30(2): 186-194. [WANG Ying, SU Yongxiu, LI Zheng. Adjusting methods for daily minimum temperature series in high altitude mountainous areas of western Guangxi [J]. Mountain Research, 2012, 30(2): 186-194] DOI: 10.16089/j.cnki.1008-2786.2012.02.013
[3] 王晶, 赵龙, 吴辉, 等. 西南地区城市化进程加剧局地气象条件的空间差异[J]. 山地学报, 2022, 40(1): 120-135. [WANG Jing, ZHAO Long, WU Hui, et al. Urbanization magnifies spatial variations of local meteorological conditions in southwest China [J]. Mountain Research, 2022, 40(1): 120-135] DOI: 10.16089/j.cnki.1008-2786.000660
[4] 李叶, 张艳红, 陈子琦, 等. 中高纬度山区气温空间化的方法比较研究——以大兴安岭北麓为例[J]. 山地学报, 2021, 39(2): 174-182. [LI Ye, ZHANG Yanhong, CHEN Ziqi, et al. Comparative study on spatialization methods of air temperature in middle and high latitude mountainous areas: A case study of northern foot of the Daxing'anling Mountains [J]. Mountain Research, 2021, 39(2): 174-182] DOI: 10.16089/j.cnki.1008-2786.000585
[5] 潘留杰, 薛春芳, 王建鹏, 等. 一个简单的格点温度预报订正方法[J]. 气象, 2017, 43(12): 1584-1593. [PAN Liujie, XUE Chunfang, WANG Jianpeng, et al. A simple grid temperature forecast correction method [J]. Meteorological Monthly, 2017, 43(12): 1584-1593] DOI: 10.7519/j.issn.1000-0526.2017.12.015
[6] 戴翼, 何娜, 付宗钰, 等. 北京智能网格温度客观预报方法(BJTM)及预报效果检验[J]. 干旱气象, 2019, 37(2): 339-344+350. [DAI Yi, HE Na, FU Zongyu, et al. Beijing intelligent grid temperature objective prediction method(BJTM)and verification of forecast result [J]. Journal of Arid Meteorology, 2019, 37(2): 339-344+350] DOI: 10.11755/j.issn.1006-7639(2019)-02-0339
[7] 赵婷婷, 高凌峰, 黄荟羽, 等. 辽阳地区智能网格产品气温预报检验订正[J]. 现代农业科技, 2023(2): 159-163. [ZHAO Tingting, GAO Lingfeng, HUANG Huiyu, et al. Test correction of intelligent grid product temperature forecast in Liaoyang area [J]. Modern Agricultural Science and Technology, 2023(2): 159-163] DOI: 10.3969/j.issn.1007-5739.2023.02.037
[8] 张成军, 纪晓玲, 马金仁, 等. 多种数值预报及其释用产品在宁夏天气预报业务中的检验评估[J]. 干旱气象, 2017, 35(1): 148-156. [ZHANG Chengjun, JI Xiaoling, MA Jinren, et al. Verification of numerical forecast and its application products in weather forecast in Ningxia [J]. Journal of Arid Meteorology, 2017, 35(1): 148-156] DOI: 10.11755/j.issn.1006-7639(2017)-01-0148
[9] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks [J]. Advances in Neural Information Processing Systems, 2012, 25(2): 1-9. DOI: 10.1145/3065386
[10] 杨绚, 代刊, 朱跃建. 深度学习技术在智能网格天气预报中的应用进展与挑战[J]. 气象学报, 2022, 80(5): 649-667. [YANG Xuan, DAI Kan, ZHU Yuejian. Progress and challenges of deep learning techniques in intelligent grid weather forecasting [J]. Acta Meteorologica Sinica, 2022, 80(5): 649-667] DOI: 10.11676/qxxb2022.051
[11] 夏景明, 戴如晨, 谈玲. 一种基于MSF-Net网络模型的短时降水预测方法:202310715521.4 [P]. 2023- 07-18. [XIA Jingming, DAI Ruchen, TAN Ling. A short-term precipitation prediction method based on MSF-Net network model: 202310715521.4 [P]. 2023- 07-18]
[12] 陈先昌. 基于卷积神经网络的深度学习算法与应用研究[D]. 杭州: 浙江工商大学, 2014: 1-22. [CHEN Xianchang. Research on algorithm and application of deep learning based on convolutional neural network [D]. Hangzhou: Zhejiang Technology and Business University, 2014: 1-22]
[13] 门晓磊, 焦瑞莉, 王鼎, 等. 基于机器学习的华北气温多模式集合预报的订正方法[J]. 气候与环境研究, 2019, 24(1): 116-124. [MEN Xiaolei, JIAO Ruili, WANG Ding, et al. A temperature correction method for multi-model ensemble forecast in north China based on machine learning [J]. Climatic and Environmental Research, 2019, 24(1): 116-124] DOI: 10.3878/j.issn.1006-9585.2018.18049
[14] 陈鹤, 蔡荣辉, 陈静静, 等. 基于深度学习方法的气温预报技术应用与评估[J]. 气象, 2022, 48(11): 1373-1383. [CHEN He, CAI Ronghui, CHEN Jingjing, et al. Application and evaluation of temperature forecast based on deep learning method [J]. Meteorological Monthly, 2022, 48(11): 1373-1383] DOI: 10.7519/j.issn,1000-0526.2002.070101
[15] 王怡, 普运伟. 基于CNN-BiLSTM-Attention融合神经网络的大气温度预测[J]. 中国水运, 2023, 23(1): 25-27. [WANG Yi, PU Yunwei. Prediction of atmospheric temperature based on CNN-BiLSTM-Attention fusion neural network [J]. China Water Transport, 2023, 23(1): 25-27]
[16] 季彦东. 基于改进LSTM模型的大气温度预测[J]. 通化师范学院学报, 2020, 41(8): 82-86. [JI Yandong. Atmospheric temperature prediction based on improved LSTM model [J]. Journal of Tonghua Normal University, 2020, 41(8): 82-86] DOI: 10.13877/j.cnki.cn22-1284.2020.08.015
[17] 王彦卷. 基于CNN-LSTM和气象要素关联的气温时空预测[D]. 银川: 宁夏大学, 2022: 1-60. [WANG Yanjuan. Spatiotemporal temperature prediction based on CNN-LSTM and meteorological elements correlation [D]. Yinchuan: Ningxia University, 2022: 1-60]
[18] 雷蕾, 徐邦琪, 高庆九, 等. 基于卷积神经网络的长江流域夏季日最高温度延伸期预报方法研究[J]. 大气科学学报, 2022, 45(6): 835-849.[LEI Lei, XU Bangqi, GAO Qingjiu, et al. Extended-range forecasting method of summer daily maximum temperature in the Yangtze river basin based on convolutional neural network [J]. Transactions of Atmospheric Sciences, 2022, 45(6): 835-849] DOI: 10.13878/j.cnki.dqkxxb.20211101001
[19] 施恩, 李骞, 顾大权, 等. 基于局部特征的卷积神经网络模型[J]. 计算机工程, 2018, 44(2): 282-286. [SHI En, LI Qian, GU Daquan, et al. Convolutional neural network model based on local features [J]. Computer Engineering, 2018, 44(2): 282-286] DOI: 10.3969/j.issn.1000-3428.2018.02.048
[20]马司周. 基于深度学习的多序列气温预测研究[D]. 兰州: 兰州理工大学, 2022: 34-51. [MA Sizhou. Research on multi-series temperature prediction based on deep learning [D]. Lanzhou: Lanzhou University of Technology, 2022: 34-51] DOI: 10.27206/d.cnki.ggsgu.2022.000500
[21] 崔海霞, 刘娜. 甘肃省旅游景点温度预报质量检验评估[J]. 甘肃科技, 2021, 37(5): 83-85+101. [CUI Haixia, LIU Na. Quality inspection and evaluation of temperature forecast for tourist attractions in Gansu [J]. Gansu Science and Technology, 2021, 37(5): 83-85+101]
[22] 王飞飞. 基于改进卷积神经网络算法的研究与应用[D]. 南京: 南京邮电大学, 2016: 10-14. [WANG Feifei. Research and applications based the improved convolutional neural network [D]. Nanjing: Nanjing University of Posts and Telecommunications, 2016: 10-14]

备注/Memo

备注/Memo:
收稿日期(Received date): 2023-11- 09; 改回日期(Accepted date):2024- 04-10
基金项目(Foundation item): 金华市科学技术局计划项目(2022-4-082)。[Jinhua Science and Technology Bureau Project(2022-4-082)]
作者简介(Biography): 郜庆林(1977-),男,浙江金华人,本科,高级工程师,主要研究方向:气象信息技术、大气探测、气象算法。[GAO Qinglin(1977-), male, born in Jinhua, Zhejiang province, B.S., senior engineer, research on meteorological information technology, atmospheric detection and meteorological algorithm]E-mail: 26250447@qq.com
更新日期/Last Update: 2024-03-30