概率预测方法在小区域人口预测中的应用——以上海市青浦区为例

Application of Probabilistic Approach to Population Forecasting in Small Areas:Based on Qingpu District,Shanghai

文章在探讨中国小区域（市县）使用概率人口预测方法的必要性和可行性的基础上，以上海市青浦区为例进行实证应用。小区域人口预测往往面临数据缺乏、数据质量不高和频繁的人口迁移流动等问题，这些都增大了人口预测的不确定性，对不确定性的科学把握是概率预测方法的优势所在。文章通过比较各种概率预测方法的适用性后，发现由Alho等（2005）发展的随机人口预测方法能够克服小区域人口预测面临的问题，从而得出科学合理的预测结果。文章使用Aoho等的方法对上海市青浦区2011～2030年的人口发展进行随机预测。

This paper addresses the application of the probabilistic approach to population forecasting in the small areas(city and town) and uses the case of Qingpu district in Shanghai to carry out empirical application.Data scarcity,poor data quality and frequent population migration and floating are the major problems in the population forecasting in small areas,which increase the uncertainty of population forecasting.The advantage of probabilistic forecasting method is to scientifically tackle such uncertainties.By comparing the different probabilistic approaches of forecasting,it finds that the stochastic population forecasting method developed by Alho and Spencer(2005) is able to overcome these difficulties and get reasonable results.Therefore,the paper forecasts population development from 2011 to 2030 of Qingpu District in Shanghai by using the stochastic population forecasting method.