1:45 PM - 3:15 PM
[O11-P100] The Current Status and Causes of Shoreline Changes Due to Coastal Erosion in Zushi
Keywords:Sea, Environment
The Zushi Coast has reportedly been experiencing erosion due to the construction of the Hayama Port A Breakwater, which was built in 2001. As a countermeasure, sand recycling is frequently conducted to transport sand from the east beach—where sediment accumulates at the river mouth—to the west beach, where erosion is more pronounced.
In order to clarify the extent of the erosion and determine whether the breakwater has had an impact, we conducted an analysis using two methods, focusing on changes in the shoreline observed through historical aerial photographs. A total of 53 aerial images, taken between 1943 and 2024 and made available by from the Geospatial Information Authority of Japan (GSI) and Google Earth, were used for this analysis. Since there is a span of more than 80 years between the oldest and newest photos, and the surrounding structures have changed significantly, we identified seven fixed points that remained unchanged throughout the years. Two of these points were used for scaling, while the remaining five—denoted as P_1 through P_5 from west to east—were used to measure the width of the coastline.
The beach width was defined as the perpendicular distance from each fixed point to the shoreline, measured along a straight line intersecting the shoreline. Each photo was corrected to a consistent scale using the two scaling points.
Using the corrected data, we examined the impact of the breakwater construction using two approaches:
Method 1 involved plotting distance on the vertical axis against time on the horizontal axis and fitting regression lines before and after 2001. If the slopes differed significantly, this would suggest an impact around the time of the breakwater's construction. However, when plotting the data for each point, the results were highly variable, and regression analysis could not be reliably performed. One likely reason for this variation is tidal influence. As a pocket beach, Zushi has a relatively narrow width—less than 100 meters even at its widest points—making it highly sensitive to tidal fluctuations. Additionally, tidal data from earlier years, as published by the Japan Meteorological Agency, was incomplete and not suitable for accurate correction.
Method 2 focused on statistical testing. For each point and the combined average of all points, we tested whether the mean beach width after 2001 was significantly less than the mean before 2001. Let n_before and n_after denote the sample sizes before and after 2001, AVE_before and AVE_after the sample means, and S_before^2 and S_after^2 the unbiased variances for each period. The null hypothesis was that there is no difference in mean width before and after 2001; the alternative hypothesis was that the mean before 2001 was greater. A right-tailed t-test was conducted using the following statistic: t=\frac{ AVE_{before} – AVE_{after} }{ \sqrt{ ( \frac{1}{n_{before}} + \frac{1}{n_{after}} )( \frac{ (n_{before} - 1)S_{before}^2 + (n_{after} - 1)S_{after}^2 }{ n_{before} +n_{after} - 2 } ) } }. The resulting p-values for P_1, P_3, and the total average were 0.016, 0.002, and 0.09, respectively—indicating statistical significance at the 5% level for P_1 and P_3, and at the 10% level for the overall average. At points P_2, P_4, and P_5, however, the post-2001 average width was greater than the pre-2001 average. These results suggest that the coastline has partially narrowed over time, particularly at some locations.
Two key issues remain. First, the number of data points before 2001 is smaller than that after 2001, so increasing the pre-2001 sample size is necessary. Second, in addition to the breakwater construction, the construction of National Route 134 may also have contributed to significant topographic changes within Zushi Bay.
In order to study the impacts and mechanisms of erosion, I wanted to use past seafloor topography data. However, since such data could not be found, a survey of the current seafloor was conducted. The survey was performed by lowering a rope with memory markers from a boat and measuring the water depths at approximately 120 locations in a grid pattern, spaced 100 meters apart, within Zushi Bay. The seafloor data was then digitized and visualized in 3D using the Python library matplotlib and the online graphing tool Desmos. By doing this, we were able to understand the current seafloor topography, and we plan to use this data for future analyses of erosion impacts and numerical simulations.
In order to clarify the extent of the erosion and determine whether the breakwater has had an impact, we conducted an analysis using two methods, focusing on changes in the shoreline observed through historical aerial photographs. A total of 53 aerial images, taken between 1943 and 2024 and made available by from the Geospatial Information Authority of Japan (GSI) and Google Earth, were used for this analysis. Since there is a span of more than 80 years between the oldest and newest photos, and the surrounding structures have changed significantly, we identified seven fixed points that remained unchanged throughout the years. Two of these points were used for scaling, while the remaining five—denoted as P_1 through P_5 from west to east—were used to measure the width of the coastline.
The beach width was defined as the perpendicular distance from each fixed point to the shoreline, measured along a straight line intersecting the shoreline. Each photo was corrected to a consistent scale using the two scaling points.
Using the corrected data, we examined the impact of the breakwater construction using two approaches:
Method 1 involved plotting distance on the vertical axis against time on the horizontal axis and fitting regression lines before and after 2001. If the slopes differed significantly, this would suggest an impact around the time of the breakwater's construction. However, when plotting the data for each point, the results were highly variable, and regression analysis could not be reliably performed. One likely reason for this variation is tidal influence. As a pocket beach, Zushi has a relatively narrow width—less than 100 meters even at its widest points—making it highly sensitive to tidal fluctuations. Additionally, tidal data from earlier years, as published by the Japan Meteorological Agency, was incomplete and not suitable for accurate correction.
Method 2 focused on statistical testing. For each point and the combined average of all points, we tested whether the mean beach width after 2001 was significantly less than the mean before 2001. Let n_before and n_after denote the sample sizes before and after 2001, AVE_before and AVE_after the sample means, and S_before^2 and S_after^2 the unbiased variances for each period. The null hypothesis was that there is no difference in mean width before and after 2001; the alternative hypothesis was that the mean before 2001 was greater. A right-tailed t-test was conducted using the following statistic: t=\frac{ AVE_{before} – AVE_{after} }{ \sqrt{ ( \frac{1}{n_{before}} + \frac{1}{n_{after}} )( \frac{ (n_{before} - 1)S_{before}^2 + (n_{after} - 1)S_{after}^2 }{ n_{before} +n_{after} - 2 } ) } }. The resulting p-values for P_1, P_3, and the total average were 0.016, 0.002, and 0.09, respectively—indicating statistical significance at the 5% level for P_1 and P_3, and at the 10% level for the overall average. At points P_2, P_4, and P_5, however, the post-2001 average width was greater than the pre-2001 average. These results suggest that the coastline has partially narrowed over time, particularly at some locations.
Two key issues remain. First, the number of data points before 2001 is smaller than that after 2001, so increasing the pre-2001 sample size is necessary. Second, in addition to the breakwater construction, the construction of National Route 134 may also have contributed to significant topographic changes within Zushi Bay.
In order to study the impacts and mechanisms of erosion, I wanted to use past seafloor topography data. However, since such data could not be found, a survey of the current seafloor was conducted. The survey was performed by lowering a rope with memory markers from a boat and measuring the water depths at approximately 120 locations in a grid pattern, spaced 100 meters apart, within Zushi Bay. The seafloor data was then digitized and visualized in 3D using the Python library matplotlib and the online graphing tool Desmos. By doing this, we were able to understand the current seafloor topography, and we plan to use this data for future analyses of erosion impacts and numerical simulations.