城市分类对中国城市化影响评估的不确定性分析

Analysis of the effect of uncertainty in urban and rural classification on urbanization impact assessment

采用美国国防气象卫星计划的夜间灯光数据DMSP/OLS、MODIS土地覆盖数据、人口数据进行城、乡气象站点的划分,探讨了三种不同划分方法对城市化不确定性的影响,并根据三种方法认定的城、乡站点,使用均一化订正的地面观测气温数据,分析了1961—2004年我国不同区域和12个城市群的城市热岛效应(Urban Heat Island effect,UHI effect)对气温变化的影响。结果表明:东南、西南和东北地区城市化影响研究的不确定性主要体现在UHI effect的数值大小上,而西北地区的不确定性则多体现在变化趋势的方向上;用三种方法同时认定的城、乡站点对比表明,中国东南地区城市化对区域气温增暖的贡献较强,西北地区较弱,东北和西南地区则无明显一致的影响;城市化对季节平均温度的影响存在南北差异,春、秋季在南方的影响显著,冬、春季在北方的影响显著。

Considerable differences in the influence of urbanization on climate change are found owing to the different methods employed to classify urban and rural observation stations.In this study, based on DMSP/OLS （Defense Meteorological Satellite program/Operational Linescan System） night light data, MODIS land cover data and population data, three different methods are used to distinguish between urban and rural observation station, from which we try to identify a reliable method of classification and study the effects of the three methods in terms of the uncertainty in urbanization impact assessment.Specifically, we analyze the urban heat island effect on changes in maximum temperature, minimum temperature and average temperature over different regions and 12 city groups in China for the period 1961--2004.The data used in this paper are air temperature data,provided by the National Meteorological Information Center of China, based on approximately 740 stations.The results show that the differ- ent classification methods yield considerable differences in the urban heat island effect, and with different regional characteristics.The uncertainties of the urban heat island effect over the southeast, southwest and northeast are mainly characterized by the absolute numerical quantities, whereas over the northwest it is characterized by the change trend.The urban heat island effect has a more significant influence on minimum temperature warming than maximum temperature and average temperature.There is a significant contribution of the urban heat island effect over the southeast, whereas the effect over the northwest is relatively weak.There is no influence of the urban heat island effect over the northeast and southwest.The urban heat island effect on the seasonal variation in temperature shows large differences between the north and south of China.Specifically, it is more significant in spring and au- tumn over South China,but in winter and spring over North China.Meanwhile, the results of the 12 city groups in China are similar to these characteristics. Temperature changes over most of the city groups are contributed to greatly by the urban heat island effect, especially in terms of the regional and seasonal effect. The urban heat island effect on minimum temperature is strongest, and the effect is more obvious over Southeast China;the value of the urban heat island effect on minimum temperature is 0. 089 1 ~C ~ （ 10 yr）-I and the contribution rate of warming is 34. 11%.Meanwhile, the urban heat island effect on maximum temperature is weakest.Although the urban heat island effect is not obvious for some of the city groups, the influence is significant in terms of the sea- sonal change.Taking the Chengdu-Chongqing region in spring as an example,the urban heat island effect on max- imum temperature can reach 0.036 4 ℃ （ 10 yr）-1, and the effect on the seasonal change in maximum tempera- ture is more significant than on minimum and average temperature.