2020, Vol.20, No.2

Journal of the Korean Cartographic Association 2023 KCI Impact Factor : 1.34

pISSN : 1598-6160
https://journal.kci.go.kr/jkca

It has been over 50 years since South Korea began to build the expressway. However, no research has been conducted on the main facilities of the expressway itself. I analyzed the distribution and name characteristics of expressway junctions which are considered important among main facilities by drivers. As of June 2020, the analysis results of 108 junctions are as follows. The simple total number of junctions for each route was higher for the main trunk lines, but the number of junctions per unit distance was much higher for shorter lines. By region, junctions were concentrated around major cities such as Gyeonggi and Gyeongnam, but unlike the size of area, the number of junctions relative to the population was smaller on the contrary. In word type analysis, ‘Sino-Korea’ word was 99.1%. In syllable analysis, the number of syllable was averaged 2.17 and syllable of junction names has exponential distribution. In name source analysis, junction name related to administrative district name was outstanding with 87.0%; junction names related to town･township name and city･county･district name accounted for the major of that. These characteristics are largely influenced by ‘the criteria for granting facility names’ that follows the administrative district name first.

This research investigates the impacts of artificial lights intensity, emitted from built-up area, on population estimation utilizing SNPP VIIRS DNB and land use and land cover (LULC) data. To do so, the performance comparison of population prediction for individual metropolitan cities and provinces in South Korea was made with simple regression model (SRM) consisting of two kinds of nighttime lights (NTL) intensity derived from only built-up area and all types of LULC. The SRM of urban land use was found to be superior to that of all LULC types in terms of positive correlation, model fit, statistical significance, and population prediction results. Nonetheless, according to regression diagnotics, this model was hard to satisfy the basic assumption of linearity. Therefore, it was required to refine the current model with either the adoption of polynomial or the transformation of variables to predict population in future research.

This study tried to quantitatively reveal influence of land cover and vegetation on DSM generation with statistical analysis. The forest played the most significant role in increase of elevation difference between DSM and DEM and the broadleaf tree forest exerted more influences on the difference than the needle leaf tree forest and mixed stand forest did, although there were slight variations depending on the tree type. This study suggests that low statistical relevance can be attributed to shooting problem and distortion of aerial photo, and problem associated with change in shape or location during shooting as well as shadow region. Despite of various causes to decrease statistical relevance, it can be suggested that results in this study directly reflected morphological characteristics of canopy by vegetation. With more data from this study, it is expected that characteristics of DSM and DEM can help to extract information such as forest attribute and tree type distribution.