人工神经网络在南海近海面气温反演中的应用研究

doi:10.11978/j.issn.1009-5470.2012.02.002

热带海洋学报 ›› 2012, Vol. 31 ›› Issue (2): 7-14.doi: 10.11978/j.issn.1009-5470.2012.02.002cstr: 32234.14.j.issn.1009-5470.2012.02.002

人工神经网络在南海近海面气温反演中的应用研究

吴新荣 ^1,2,3,4 , 韩桂军 ^1,2 , 张学峰 ^1,2,3,4 , 王喜冬 ^1,2,3,4

1. 国家海洋局海洋环境信息保障技术重点实验室 , 天津 300171; 2. 国家海洋信息中心 , 天津 300171; 3. 中国科学院南海海洋研究所 , 广东广州 510301; 4. 中国科学院研究生院 , 北京 100039

收稿日期:2010-11-10 修回日期:2011-01-18 出版日期:2012-06-05 发布日期:2012-06-05
作者简介: 吴新荣 (1981 — ), 男 , 江苏省泰州市人 , 在读博士生 , 主要从事海洋数据同化方法应用研究。
基金资助:
国家自然科学 (41030854 、 40906015 、 40906016); 国家科技支撑计划项目 (2011BAC03B 02-01-04 )

Retrieving near-surface air temperature in the South China Sea using artificial neural network

WU Xin-rong^1,2,3,4, HAN Gui-jun^1,2, ZHANG Xue-feng^1,2,3,4, WANG Xi-dong^1,2,3,4

1. Key Laboratory of Marine Environmental Information Technology, SOA, Tianjin 300171, China 2. National Marine Data and Information Service, Tianjin 300171, China 3. South China Sea Institute of Oceanology, Chinese Academy of Sciences, Guangzhou 510301, China 4. Graduate University of Chinese Academy of Sciences, Beijing 100039, China

Received:2010-11-10 Revised:2011-01-18 Online:2012-06-05 Published:2012-06-05

摘要/Abstract

摘要： 基于人工神经网络方法 , 利用海面水温、海面风速以及海面气压反演南海近海面气温 , 采用的基础数据集是国际综合海洋 - 大气数据集 (International Comprehensive Ocean-Atmosphere Data Set, 2.4 Release, ICOADS2.4)1981 — 2008 年的观测资料 , 其中 1981 — 2000 年的观测资料用来建立模型 , 2001 — 2008 年的观测资料用来进行模型检验。采用的人工神经网络方法是引入动量因子并采用批处理梯度下降法的 BP(Back propagation) 算法。试验结果表明 , 基于人工神经网络建立的近海面气温反演方法明显优于多元线性回归方法 , 尤其是在春季和冬季 , 海面水温、海面风速以及海面气压与近海面气温之间存在较强的非线性关系 , 人工神经网络的优势更加明显。总体而言 , 人工神经网络在各月的反演效果较均衡 , 均方根误差介于 1.5— 1.8 ℃ 之间 , 平均绝对误差为 1.1— 1.3 ℃ 。

关键词: 人工神经网络, BP算法, 多元线性回归

Abstract: Based on artificial neural network (ANN), the authors retrieved near-surface air temperature (AT) from sea surface temperature (SST), wind speed (WS) and sea level pressure (SLP) of the International Comprehensive Ocean-Atmosphere Dataset (ICOADS). Modeling sample spans from 1981 to 2000, while validating sample spans from 2001 to 2008. The adopted ANN introduces momentum factor to back propagation (BP) algorithm to escape from local extremes. In addition, batch processing gradient descent method was used to remove the effect of sequential training. Retrieving results in the South China Sea (SCS) demonstrates that ANN is better than multi-factor linear regression, especially for coastal areas during spring and winter, where strong non-linear relation exists between SST, WS, SLP and AT. In conclusion, ANN behaves similarly for each month, with root mean square error (RMSE) between 1.5 ℃ and 1.8 ℃ and mean absolute error (MAE) between 1.1 ℃ and 1.3 ℃ .

Key words: Artificial neural network, BP algorithm, Multi-factor linear regression

中图分类号:

P731

吴新荣,韩桂军,张学峰,王喜冬. 人工神经网络在南海近海面气温反演中的应用研究[J]. 热带海洋学报, 2012, 31(2): 7-14.

WU Xin-rong,HAN Gui-jun,ZHANG Xue-feng,WANG Xi-dong. Retrieving near-surface air temperature in the South China Sea using artificial neural network[J]. Journal of Tropical Oceanography, 2012, 31(2): 7-14.

参考文献

KUBOTA M, SHIKAUCHI A. Air temperature at ocean surface derived from surface-level humidity[J]. J Oceanogr, 1995, 51: 619-634. KONDA M, IMASATO N, SHIBATA A. A new method to determine near-sea surface air temperature by using satellite data[J]. J Geophys Res, 1996, 102(c6): 14349-14360. JONES C, PETERSON P, GAUTIER C. A new method for deriving ocean surface specific humidity and air temperature: An artificial neural network approach[J]. J App Meteorol, 1999, 38: 1229-1245. SINGH R, VASUDEVAN B G., PAL P K et al. Artificial neural network approach for estimation of surface specific humidity and air temperature using multifrequency scanning microwave radiometer[J]. J Earth Syst Sci, 2004, 113(1): 89-101. SINGH R, KISHTAWAL C M, JOSHI P C. Estimation of monthly mean air-sea temperature difference from satellite observations using genetic algorithm[J]. Geophys Res Lett, 2005, 32, L02807.1-L02807.5. SINGH R, JOSHI P C, KISHTAWAL C M. A new method to determine near surface air temperature from satellite observations[J]. Int J Remote Sens, 2006, 27: 2831-2846. JACKSON D L, WICK G A, BATES J J. Near-surface retrieval of air temperature and specific humidity using multisensor microwave satellite observations[J]. J Geophys Res, 2006, 111, D10306, doi:10.1029/2005JD006431 伍玉梅, 何宜军, 孟雷. 利用卫星资料反演月平均近海面气温和湿度[J]. 海洋与湖沼, 2008, 39(6): 546-551. 王丽静, 何宜军, 张彪: 利用遗传算法反演实时近海面气温和比湿[J]. 海洋科学, 2009, 33(4): 20-24. 蒋宗礼. 人工神经网络导论[M]. 北京: 高等教育出版社, 2001: 39-54. ZHUANG W, XIE S P, WANG D X, et al. Intraseasonal variability in sea surface height over the South China Sea[J]. J Geophys Res, 115, C04010, doi:10.1029/2009JC005647. 陈颖珺, 谢强, 蒙伟光, 等. 不同海表面温度对南海台风“杜鹃”的影响试验[J]. 热带气象学报, 2009, 25(4): 401-406. QIU C H, WANG D X, HIROSHI K, et al. Validation of AVHRR and TMI-derived sea surface temperature in the northern South China Sea[J]. Cont Shelf Res, 2009, 29: 2358-2366. SHU Y Q, ZHU J, XIAO X J, et al. Performance of four sea surface temperature assimilation schemes in the South China Sea[J]. Cont Shelf Res, 2009, 29: 1489-1501.

人工神经网络在南海近海面气温反演中的应用研究

Retrieving near-surface air temperature in the South China Sea using artificial neural network

PDF (PC)

赞

可视化

被引次数

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 1

Metrics

本文评价

推荐阅读 0