Based on the automatic fare collectionc （AFC）， data of Changsha Metro， intercity tourists who arrived at Changsha via the metro external transportation transfer stations are identified. Based on their stay time in Changsha， these tourists are clustered into three groups using Gaussian mixture model， and they are mainly identified to three stay pattern types： the day-return type， the one-night type and the two-night type. The research results show that in terms of travel purpose， all three groups mainly travel to the Wuyi Square-centered core business district of Changsha. Generally， the proportion of overnight passenger flow is higher when it is far away from the city center， and the proportion of the day-return passenger flow increases when it is close to the city center. The day-return passenger flow mostly meets short-term travel needs， such as sightseeing and business. The two-night group has the more complex travel needs， such as medical treatment and transfer. The one-night group is a transitional group that has both characteristics. In terms of travel time， the group that stays in Changsha for a longer time usually has a longer metro travel time. The study of travel characteristics of intercity tourist passenger flow can provide decision-making references for relevant transportation planning and management departments， help optimize the rail transit network， and improve the travel experience of passengers.