寒潮天气下合肥市主城区微博时空特征分析

Spatial and Temporal Characteristics of Micro-blog in Main Urban Area of Hefei under Cold Wave

传统气象灾害观测数据存在数据稀疏、滞后以及高成本等特性,利用微博等社会感知数据的高分辨率特性,可以分析灾害性天气条件下群体公共行为与用户地理位置的相关性。该文利用自主设计的一种突发灾害天气微博数据采集系统,获取合肥市2016年1月19日到1月27日的寒潮天气下新浪微博数据,利用GIS空间分析等方法探索寒潮发生前、中、后的微博用户时空聚集特征。研究发现,在时间上,微博发布量的变化与寒潮发生存在一定的相关性,一天当中18时至24时是微博高发时段;在空间上,微博发布地主要集中在一环二环内的商业聚集区、交通集散地、旅游观光中心与高校园区等人流量密集区域。在寒潮词频统计方面,"雪""寒""冷"和"冻"等一系列与寒潮相关的热词,在寒潮期间出现频率有突升的趋势。该研究有助于气象灾害应急响应与决策。

Due to observations from meteorological typically suffering from data sparsity,time delays and high costs,we can analyze the correlation between the group public behavior patterns and geographical locations under disastrous weather by using rapid high-resolution observations of social sensing,such as micro-blog data.This paper designed and developed a sudden disaster weather data acquisition system for micro-blog,taking the data of SINA Microblog under cold wave weather,during a period of8days in2016started on January19th and ended on January26,in main urban area of Hefei as an example,and explored the spatio-temporal aggregation characteristics of micro-blog users before,during and after the cold wave based on GIS spatial analysis methods.The study showed that,in time,the certain correlation between the changed messages of micro-blog and the cold wave was present during the period of6hours started18p.m.and ended24p.m.Also,the period was the high-frequency time of micro-blog throughout the day.In space,the micro-blog messages mainly concentrated the business districts,traffic distribution centers,tourist centers,campus cities and other areas of the densely populated area.In the word frequency statistics of cold wave,a series of hot words,such as"cold","frozen"snow,etc.increased sharply during the cold period.The study could assist the emergency response and decision-making of meteorological disasters.