京津冀地区AOD时空变化及影响因子的地理探测

基于MODIS 3 km AOD遥感数据,利用空间自相关模型及地理探测器对2010—2016年京津冀地区AOD的时空变化特征及其影响因子进行探测。结果表明:(1) 2010—2016年京津冀地区年平均AOD值为0.83,其中天津市年均AOD值为研究区最高,河北省次之,北京市最低。研究区及各分区AOD的7 a变化趋势大体一致,呈现先下降后上升再小幅波动的状态。(2)空间自相关分析表明京津冀地区AOD空间分布呈现显著正相关。局部高高聚集区主要集中在北京市东南部、天津市南部及河北省的中南部,低低聚集区集中分布在西北部山区。研究区高低聚集区面积均呈减小趋势,不显著区呈扩大态势。(3)地理探测器结果表明不同区域的主导影响因子不同,北京市首要影响因子为NDVI,其次为人口密度,且二者交互作用明显。天津市主导因子为风速,人口密度、第二产业生产总值等人为因子的作用力也较大,风速与其交互作用较强。河北省主导因子为人口密度,GDP与第二产业生产总值等的作用力次之,整体交互作用偏弱。通过地理探测器解析京津冀地区AOD空间分异的影响机理,其结果对我国大气污染治理具有重要意义。

Spatiotemporal variations of AOD and geographical detection of its influence factors in Beijing-Tianjin-Hebei region

TheBeijing-Tianjin-Hebei region, which is located in the Bohai Rim, is the mosteconomically developed region in northern China. It is also one of the mostpolluted regions in China. Aerosol optical depth (AOD) can effectively reflect the degree of air pollution. In thisstudy, the MODIS 3 km AOD remotesensing data was analyzed and the research results would be of greatsignificance to China’s air pollution. Firstly, the temporal variation characteristicsand spatial autocorrelation characteristics of AOD in Beijing-Tianjin-Hebei region from 2010 to 2016 wereanalyzed. The factors of precipitation, wind speed, relative humidity, normalizeddifference vegetation index (NDVI),grossdomestic product (GDP),secondary industry GDP, population density were selectedas the affecting factors of AOD, usinggeographical detection to analyze the trend of the factors contribution ratioand its dominant affecting factors in different regions and different times inBeijing-Tianjin-Hebei region. The results showed as follows: (1) The annualaverage AOD of the Beijing-Tianjin-Hebeiregion was 0.83 from 2010 to 2016.The annual average AOD of Tianjin was the highest in the study area, followed by Hebeiand Beijing. The annual variation trend of AOD was shown an overall state of decline first followed by rise and then a smallfluctuation change and a similar status in each local region as well. (2) Thespatial autocorrelation analysis indicated a significant positive correlationamong spatial distribution of AOD inBeijing-Tianjin-Hebei region. Local high-accumulation areas were mainlyconcentrated in the southeastern part of Beijing, southern part of Tianjin, andcentral-southern part of Hebei Province. The low-accumulation areas wereconcentrated in the mountains of the northwest. The area of the high and lowaccumulation areas in Beijing-Tianjin-Hebei region showed a decreasing trend, whilethe area of the non-significant areas showed an increasing trend. (3)Geographic detection analysis indicated that the primary influencing factor was NDVI in Beijing, followed bypopulation density, and their interaction was significant. The dominant factorwas wind speed in Tianjin. The effect of human factors such as populationdensity and the GDP of the secondary industry were also important. The effectof interactions between wind speed and the above factors were also significant.The dominant factor was population density in Hebei Province, followed by GDPand the secondary industry’s GDP, and the overall interaction was relativelyweak. In the past, the correlation analysis of the influencing factors wasbased on linear relationship of the overall area, while the geographicaldetection not only has wireless assumptions, but also reveals the linear, nonlinearand spatial relationships of the driving factors. From this perspective, thegeographical detection is more reliable in studying the influencing factors. Thisstudy applied geographical detection to analysis of the impacting factors of AOD in Beijing-Tianjin-Hebei region, whichproved its feasibility in this field and this region. At the same time, theresults also had important reference value for air pollution prevention, industrialand agricultural layout and urban construction in Beijing-Tianjin-Hebei region.