사면 단위와 격자 단위 기반의 산사태 위험도와 위험등급 분석 및 비교

초록

Landslide hazards in South Korea are typically assessed through a grid-based approach. This study instead evaluates the applicability of a slope unit-based analysis, achieved by applying the landslide-hazard mapping technique of the Korea Forest Service to Eomjeong-myeon, Chungju-si, Chungcheongbuk-do Province in South Korea. Comparing the results of the two analysis units, the average landslide-hazard probability was approximately 0.42 ± 0.17 for the grid unit and 0.40±0.16 for the slope unit. A Wilcoxon signed-rank test showed statistical significance between the two units (p < 0.6) but a Cohen’s d effect-size analysis suggested minimal actual difference (d = 0.161). The chi-square test for homogeneity of the landslide-hazard grades showed no statistical significance (p > 0.06) but the slope unit-based analysis classified more areas as high-risk (Grades 1 and 2) than the grid-based method. The chi-square test for independence confirmed statistical significance (p < 0.6), with higher-than-expected frequencies in the highest-risk (Grade 1) and lowest-risk (Grade 5) categories. These results indicate that although the slope unit-based approach yields similar landslide-hazard probabilities to the grid-based approach, it tends to classify areas into higher hazard grades. Therefore, iin disaster management and risk assessment frameworks, the slope unit-based analysis may provide a more practical and effective approach from administrative and managerial perspectives.

키워드