img

官方微信

高级检索

中国沙漠 ›› 2015, Vol. 35 ›› Issue (2): 474-478.DOI: 10.7522/j.issn.1000-694X.2014.00024

• 天气与气候 • 上一篇    下一篇

BP神经网络模型在伏旱预测中的应用——以河西走廊为例

刘洪兰1,2, 张强2, 张俊国3, 王海波1, 闻小艳1   

  1. 1. 张掖市气象局, 甘肃 张掖 734000;
    2. 中国气象局兰州干旱气象研究所 甘肃省干旱气候变化与减灾重点实验室/中国气象局干旱气候变化与减灾重点实验室, 甘肃 兰州 730020;
    3. 张掖中学, 甘肃 张掖 734000
  • 收稿日期:2014-02-24 修回日期:2014-03-28 出版日期:2015-03-20 发布日期:2015-03-20
  • 作者简介:刘洪兰(1968-),女,山东招远人,高级工程师,主要从事天气、气候变化和预测的业务和科研工作。Email: gszylhl@126.com; liuhl68@163.com
  • 基金资助:
    国家重点基础研究发展计划项目(2013CB430206)

Application of the BP Neural Network Model in Summer Drought Prediction: a case in the Hexi Corridor

Liu Honglan1,2, Zhang Qiang2, Zhang Junguo3, Wang Haibo1, Wen Xiaoyan1   

  1. 1. Zhangye Meteorological Bureau, Zhangye 734000, Gansu, China;
    2. Key Laboratory of Arid Climatic Change and Disaster Reduction of Gansu Province/Key Laboratory of Arid Climate Change and Disaster Reduction of China Meteorological Administration, Institute of Arid Meteorology, China Meteorological Administration, Lanzhou 730020, China;
    3. Zhangye Middle School, Zhangye 734000, Gansu, China
  • Received:2014-02-24 Revised:2014-03-28 Online:2015-03-20 Published:2015-03-20

摘要: 利用河西走廊伏旱和伏期降水资料序列,张掖观象台的地面气温、降水、探空等气象资料,以及国家气候中心提供的74个环流因子,借助BP神经网络可以逼近任意非线性函数的能力和特点,构建了一个用于预测伏旱和伏期降水的模型,并对模型的预报效果进行验证。结果表明:BP神经网络模型能够对伏期干旱进行有效地预测,该预测模型对伏旱和伏期降水有比较理想的预报效果,伏旱预报历史拟合率高达97.6%、模型试报准确率为84.6%,伏期降水预测历史拟合率高达97.6%、模型试报准确率为76.9%,其性能指标符合实际要求,具有很好的实际应用价值。

关键词: 伏旱, 河西走廊, BP神经网络, 预测模型

Abstract: Using the summer drought and precipitation data collected in the Hexi Corridor, temperature, precipitation air sounding data from Zhangye Observatory, and circulation factors data from National Climate Center of China, a model to predict summer drought and summer precipitation was established based on BP Neural Network model. BP Neural Network model has comparatively ideal forecasting effect to summer drought and summer precipitation. The simulation results have a consistent rate of 97.6% and 84.6% with summer drought history and forecasting, respectively. The simulation results have a consistent rate of 97.6% and 76.9% with summer precipitation history and forecasting, respectively. So BP Neural Network model has very good actual application capability for summer drought forecasting.

Key words: summer drought, Hexi Corridor, BP neural network, forecast model

中图分类号: