The spatial-temporal distribution of potential source of tourists is an important knowledge for authorities in tourist destinations. Early investigations on such issue rely on field surveys which require extensive manpower, whilst continuity and consistency of the data cannot be ensured. Hence recent approaches exploit location based services data to accurately and automatically identify the sources of tourists. This study extends such researches by introducing regional accessibility that is acquired through analyzing travel route data from internet sources. Specifically, we use Tencent LBS data to extract air, train and bus traffic flow volumes towards Huangshan, a typical tourism city, during public holidays. The volumes are then correlated with distance, train or flight shifts and other accessibility indicators to find patterns in how transportation cost and opportunity influences tourist attraction. Knowledge as such can be applied to optimizing tourist service and formulating city propaganda strategies.