그린투어리즘 포텐셜 분석을 위한 관광마을 수준의 월별 방문객 추정 - 하회마을을 중심으로 -

Abstract

Summary and Conclusion In this study, ARIMA forecasting method which has not been widely applied to rural green tourism was used to build a general model and determine the optimal length of time series for village-based monthly tourist arrival forecasting and pattern analysis. The data gotten from Hahoe Village Superintendent's Office were monthly tourist arrival data that are significant rare data for tourism villages in Korea. Using four different lengths of time series and ten alternative models for each time series, this study found that 6-year time series with the highest R 2 values is the best length of time series and SARIMA(2,1,1)(1,1,2)12 model which exists in all the lengths of time series was selected as the general model to forecast village-based tourist arrivals. The proposed general model was verified using three 6-year time series for model building and the rest of years' data for model verification. The results show that most of R 2 values between observed and predicted data are more than 0.7 except two. This means that SARIMA(2,1,1)(1,1,2)12 model and 6-year data are adequate for village-based monthly tourist arrival forecasting. Although the number of tourist arrivals can be easily affected by social and economical factors, forecasting tourist arrivals is still necessary. The results of this study are very useful for local governments of village-based tourism attractions, expecially those that lack monthly tourist arrival data, to make timely and reasonable tourism policy. On the one hand, it’s necessary to improve the quality and range of general infrastructure, tourism facility and service for peak tourist season. On the other hand, it’s essential to develop new tourism resources and products to promote the local off-season tourism. What is more important is sustainable tourism which needs the conservation of local environment and natural and cultural heritage.