Jinmu CHOI, Kyung Hee University, Korea (Republic of)
Suyeon HWANG, Kyung Hee University, Korea (Republic of)
Fine dust is a disaster phenomenon with spatiotemporal dependence, and it is appropriate to check the spatial pattern and time series pattern of the concentration distribution together. The Burstiness Index, an index for confirming the time series characteristics of an event, determines the persistence and irregularity by the time interval of the event occurrence [1, 2]. This study attempted to compare patterns of high concentrations of fine dust by region through a Local Indicator of Temporal Burstiness (LITB) that can confirm the spatial distribution of time series patterns [3].
Data that can confirm the occurrence of high concentrations of fine dust are fine dust warnings data which are part of Korea's fine dust prevention policy. First, to check the difference in high concentration warning occurrence according to particle size, PM10 and PM2.5 warning data were classified, and then LITB maps were produced for each year through the entire period of collection period (2015-2020). For comparison, simple warning frequency maps were also produced.
As a result of comparing the LITB map and the warning frequency map corresponding to the entire period, the difference in time series patterns that could not be confirmed through the frequency map could be confirm on the LITB map. In other words, simple frequency map could not show temporal burstiness of dust concentration.
Through this study, it can be suggested that the LITB can be used as a way to recognize the temporal pattern of certain phenomenon by region and enable efficient application of related problem-solving methods.
Mots clés : Local Indicator of Temporal Burstiness|Burstiness Index|Fine dust|Time series analysis|Spatial pattern
A103239JC