基于时间序列建模的城市热岛时间尺度成分分离方法与应用

A Method for Separating Temporal Components of the Urban Heat Island Based on Time Series Modeling and Its Application

城市热岛效应是全球与区域气候变化研究中的焦点问题。基于2001—2012年较长时间序列的北京市MODIS地表温度产品及相关NDVI和反射率产品,给出地表温度时间序列构建方法。基于站点气象观测资料进行的精度验证表明地表温度时间序列构建方法可行,并最终给出城市热岛强度的量化方案。研究选取统计学中X-11-ARIMA时间序列建模方法,分离并分析城市热岛强度时间序列的结构性成分。分析发现,以平均城乡温差为指标的北京城市热岛强度季节性特征明显,与城乡土地利用状况、季节性地表覆盖、地物热特性以及气候因子等联系密切。趋势—循环特征与城市扩张速度及入选城市区域面积相关。以已发生城市热岛区域城乡平均温差为指标的北京城市热岛强度趋势—循环特性在12年间表现平稳。时间序列建模分析提取出不规则变动成分,为定量研究偶然因素对城市热岛的影响提供了可能。

Urban heat island （UHI） effect has been the focus on the research of global and regional climate change. In this study, the Moderate Resolution Imaging Spectroradiometer （MODIS） land surface temperature products and their corresponding datasets of Normalized Difference Vegetation Index （NDVI） and Reflectance products acquired from 2001 to 2012 in Beijing were selected as the data sources to support the method of constructing surface temperature time series. Validation with in situ meteorological datasets revealed that the method of constructing surface temperature time series was applicable and feasible with high accuracy, and eventually quantization scheme of the UHI intensity was given. Statistical model X-11-ARIMA was selected to decompose and analyze the UHI time series. Results indicate that when using the average temperature difference between urban and rural areas as an in- dex, Beijing UHI intensity shows obvious seasonal characteristics, which are closely related to urban and rural land use status, seasonal surface coverage, thermal characteristics of ground objects, and climate factors. In the mean time, cycle trend features are associated with urban expansion. When using the average temperature difference between urban, where UHI has occurred, and rural areas as an index, cycle trend features have a stable performance. The irregular factors extracted by time series modeling analysis make the quantitative study on the accidental factors influence the urban heat island possible.