Estimating earthquake exposed population through mobile data and high-resolution imagery: a case study in main residential area of Zhangjiakou City, China

Exposed population is an important element that is used to estimate earthquake casualty, and the indoor rate is a key variable to calculate exposed population. Cell phone data can provide useful information in a timely manner to estimate the indoor rate. However, there is no previous study to get indoor rate with cell phone data. The objectives of this study is to calculate the indoor rate and estimate the exposed population in residential areas with the position data of mobile phones. The study area was divided into regular hexagons. The hourly number of mobile phones within each hexagon was added up and the value assigned to the hexagon area. The hourly indoor rate was gotten from max value of mobile phone numbers and hourly number in each hexagon. The building height, base area and population data was used to produce population density. The exposed population was calculated from population density and indoor rate. Results show that the daily trends of indoor rate (by hour) were similar for each day of the study period, with the highest values occurring at 07:00, 18:00, and between 21:00 and 07:00 the next day. The lowest values occurred at 13:00. As expected, the trend in population exposure followed a similar trend to that of the indoor rate. But, the spatial distribution of population exposure was various. These results were highly accurate. This indicates that this estimated method of population exposure can be used for future estimates of earthquake casualties.