资源科学 ›› 2017, Vol. 39 ›› Issue (6): 1212-1223.doi: 10.18402/resci.2017.06.20

• • 上一篇    下一篇

北京市行业经济产出对气象变化的敏感性分析

孙鉴锋1(), 王冀2, 何桂梅3, 陈志泊1, 王建新1()   

  1. 1. 北京林业大学信息学院,北京 100083
    2. 北京市气象局气候中心,北京 100089
    3. 北京市应对气候变化研究中心,北京 100031
  • 收稿日期:2016-08-07 修回日期:2017-04-01 出版日期:2017-06-20 发布日期:2017-06-20
  • 作者简介:

    作者简介:孙鉴锋,男,山东潍坊人,博士生,主要从事计量经济学和机器学习等研究。E-mail:jianfeng.sunmt@gmail.com

  • 基金资助:
    科学研究与研究生培养共建项目(2015);北京市发展与改革委员会适应气候变化信息模型建设(2015ZXKFXX001)

Sensitivity analyses of industrial economic output to weather variability in Beijing

Jianfeng SUN1(), Ji WANG2, Guimei HE3, Zhibo CHEN1, Jianxin WANG1()   

  1. 1. School of Information,Beijing Forestry University,Beijing 100083,China
    2. Climate Center,Beijing Meteorological Bureau,Beijing 100089,China
    3. Beijing Research Center for Climate Change,Beijing 100031,China
  • Received:2016-08-07 Revised:2017-04-01 Online:2017-06-20 Published:2017-06-20

摘要:

本文利用2002-2013年经济和气象的历史数据,分析了北京市各行业经济产出对气象因素变化的敏感性。通过改进Cobb-Douglas(C-D)生产函数这一计量经济模型,建立了气象因素变化和行业经济产出之间的数量因果关系;采用岭回归模型对北京市的行业经济-气象系统要素进行分析,得到了北京市各行业对气象条件的敏感性排名。即,建筑业、批发零售业和金融业对气象条件变化表现出高敏感性,而农业对气象条件变化的敏感性最低;从高到低依次为建筑业(0.4995)、批发与零售业(0.4176)、金融业(0.2933)、交通运输仓储和邮政业(0.2806)、工业(0.2799)、住宿和餐饮业(0.2710)、卫生与社会保障和社会福利业(0.2691)、文化体育和娱乐业(0.2607)、农业(0.2537)。通过度量行业经济对气象因素变化的敏感性大小,有利于北京政府理解这些影响并科学地进行产业结构调整和资源格局优化。研究结果表明,岭回归获得的结果更加符合北京市行业经济发展的实际情况。

关键词: 计量经济模型, 气象因素, 敏感性行业, 岭回归, 北京市

Abstract:

Based on 12 years(2002-2013)of economic data and historical weather observations,we analyzed the sensitivity of the industry economy to changes in meteorological factors. By improving an econometric model,Cobb-Douglas(C-D) production function,a quantitative causal relationship is established between meteorological factors and industry economy. Ridge regression modeling was employed to analyze meteorological factors and the industry economy of Beijing. Sensitivity ranking of Beijing economic industry was obtained,three among which are hypersensitive:construction industry,wholesale and retail trade,and financial industry. Agriculture is the least sensitive. The sensitivity ranking,from highest to lowest was:construction industry (0.499 5),wholesale and retail trade(0.417 6),financial industry(0.293 3),transportation,warehousing and postal services(0.280 6),industry(0.279 9),accommodation and catering industry(0.271 0),health and social security and welfare(0.269 1),cultural and sports and entertainment industry(0.260 7),and agriculture(0.253 7). Sensitivity analysis is helpful to the government of Beijing when conducting the industrial restructuring and optimization of resource patterns. This research indicates ridge regression is more in keeping with local economic development in Beijing.

Key words: econometric model, meteorological factors, sensitive industries, ridge regression, Beijing City