Abstract: The urban An urban tourism regional system consists of tourism service facilities and tourist attractions. Accommodation facilities, which are the part of tourism service facilities, represent regional progress and invest? ment opportunities. Spatial distribution patterns have an important effect on the development of the urban accommo? dation industry and urban spatial structure. In addition, it is crucial to explore the relationship between accommoda? tion facilities and tourist attractions for regional tourism systems as the spatial coupling and coordination of these two parts determine the structure, function, and level of a tourism regional system. Urumqi, a large city in western China, has seen an increasing number of tourist receptions and resources in recent years. Tourism has become a strategic pillar industry in Urumqi; however, the temporal and spatial variation of accommodation has not been ex? plored and the coupling relationship between the city and the tourism area system has hardly been studied. There? fore, this paper utilizes the kernel density estimation and standard deviational ellipse model to explore the spatial changes of accommodation facilities using Urumqi as a case. Then, the spatial coordination relationship between ac? commodation facilities and tourist attractions in Urumqi is explored based on fractal dimension theory and spatial data statistics methods. We found that the overall pattern of accommodation facilities in Urumqi changed significant? ly, increasing from 915 to 3,123 between 2012 and 2018. The accommodation facilities are basically distributed in the axial direction along important transportation lines such as Hetan highway and Altay road. Additionally, the number of accommodation facilities in the Midong New Area increased significantly, forming a concentration there. In 2018, accommodation facilities were shifting to the northeast significantly and axial aggregation model evolved in? to patterns of increasing trends of discrete distribution, with the most obvious trend being the dispersal of three-star and economic accommodations. The ordinary type of accommodation facilities in the main urban area has trans? ferred from isolated cores to a belt multi-core model. Additionally, budget hotels have changed from point aggrega? tion to a belt aggregation model. The aggregate fractal dimension results show that tourism attractions, with the Inter? national Grand Bazaar as the core, have evolved into a self-organizing fractal system. However, the fractal structure of the tourism service facility, which has the Southern Airlines Pearl International Hotel as its core, is at a low level. The integrated aggregation dimension of tourist facilities and tourist attractions is 1.18, demonstrating that the spa? tial coupling of these two parts is high. However, the fractal structure of tourist attractions and tourist service facili? ties has not reached an ideal state and must be optimized and adjusted. To further measure the spatial relationship between these two parts, this study establishing a fishnet grid by calculating the number of the two parts in each grid. These results yielded a Pearson correlation coefficient of 0.77, showing a significantly strong correlation. A lin? ear regression equation is obtained by counting the number of the two parts in a five-level buffer of the main road, and the R2 value obtained was 0.926. This demonstrates that the accommodation facilities and tourist attractions in Urumqi’s tourism regional system have strong spatial coupling and coordination, and the relevance of industrial functions determines their proximity in spatial distribution.

Key words: POI, accommodation industry, tourist attractions, aggregation fractal, Urumqi