贝叶斯层次时空模型在甘肃省AOD时空分布中的应用

Application of Bayesian Hierarchical Spatio-temporal Model in Spatio-temporal Distribution of Aerosol Optical Depth in Gansu Province

探究甘肃省大气气溶胶分布及变化特征,准确了解甘肃省气溶胶光学特性对研究大气环境污染、应对气候变化是非常重要的。基于MODIS MAIAC高分辨率数据,利用贝叶斯层次时空模型(Bayesian Hierarchical Spatio-Temporal Model,BHM)从时间趋势、空间格局和局部变化3个方面分析甘肃省气溶胶光学厚度(Aerosol Optical Depth,AOD)时空特征,结果表明:(1)从年际尺度上看,2001—2020年AOD年际变化趋势整体维持在0.4左右,2010—2013年变化趋势呈“V”形,起伏变动较大;从季节尺度上看,AOD季节变化呈现出春季整体高,夏季、秋季局部中心高,冬季水平低的变化特征。(2)从空间上看,AOD分布的空间格局为“东高西低、中部过渡”的趋势,冷、热点区域空间分布差异明显,热点区域集中分布在陇中、甘南和陇南地区。(3)局部AOD分布整体呈现以陇中、河西西部地区为中心向四周降低的趋势。

Exploring the distribution and change characteristics of atmospheric aerosols in Gansu province and accurately understanding the optical properties of aerosols in Gansu is very important for studying atmospheric pollution and coping with climate change.Based on MODIS MAIAC high resolution data,using BHM to analyze the spatial and temporal characteristics of AOD in Gansu from three aspects:time trend,spatial pattern and local changes,the results show that:(1)On the interannual scale,the interannual variation trend of AOD from 2001 to 2020 remained at around 0.4 as a whole,and the variation trend from 2010 to 2013 was in a“V”shape with large fluctuations.From the seasonal scale,the seasonal variation of AOD presents the characteristics of an overall high in spring,high local center in summer and autumn,and low level in winter.(2)From a spatial perspective,the spatial pattern of AOD distribution is a trend of“high in the east and low in the west,with a transition in the middle”.The spatial distribution of cold and hot areas is significantly different,and the hot areas are concentrated in Longzhong,Gannan and Longnan regions.(3)The overall distribution of local AOD showed a downward trend from the center of Longzhong and the west of Hexi to the surrounding areas.