Annual Journal of Hydraulic Engineering, JSCE, Vol.**, 2**, February IVESTIGATIO OF THE RIVER MOUTH MORPHOLOGY CHAGES I SAMEGAWA RIVER IDUCED BY THE 2 GREAT EAST JAPA EARTHQUAKE TSUAMI Mohammad Bagus ADITYAWA,Hitoshi TAAKA2 and Hisao AGABAYASHI3 Member of JSCE, Dr. of Eng., Dept. of Civil Engineering, Tohoku University (Aobayama, Aoba-ku, Sendai 98-8576, Japan) 2 Fellow Member of JSCE, Dr. of Eng., Professor, Dept. of Civil Engineering, Tohoku University (Aobayama, Aoba-ku, Sendai 98-8576, Japan) 3 Fellow Member of JSCE, Dr. of Eng., Professor, Dept. of Civil Engineering, ihon University (Tokusada, Koriyama, 963-8642, Japan) The river mouth morphology changes in the Samegawa river mouth during its recovery following The Great East Japan Tsunami of 2 was investigated in this study by analyzing the water level data and aerial photo. The aerial photos provide valuable information on the changes. However, their availability is limited. The correlation coefficient and the linear gradient between the water level data in the river and the tidal level were analyzed. The river mouth narrowing, as well as the river mouth opening has been successfully detected. The assessed parameters correspond well with the condition obtained from the aerial photo analysis when available. In addition, these parameters also provide information on the river mouth condition when there are no aerial photo data. The proposed method in this study will provide an efficient way for analyzing and evaluating the morphology changes at a river mouth. Key Words : Tsunami, river mouth, water level, tidal level, aerial photo. ITRODUCTIO The Great East Japan Tsunami 2 had caused severe damages to the affected coastal area. The tsunami wave height in some places may reach up to about 4 meters ). The tsunami wave has changed the coastal morphology including at the river mouth, especially in sandy area, due to the severe erosion. Some of the most affected coastal area are located in the Miyagi and Fukushima prefectures. The tsunami wave propagation on land and in river in Sendai plain, Miyagi has been reported in detail 2). In addition, the assessment of the tsunami impact in Miyagi prefecture based on aerial photo has been conducted 3). They have found that the shoreline retreated due to severe erosion occurred, especially in sandy area. The land subsidence and erosion has been reported to occur in the northern Sendai coast area 4). A more detail study on the morphology changes of the river mouth in the Miyagi and Fukushima prefectures had concluded that the river mouth with sand formation suffered greatly from the tsunami 5). The coastal morphology after the tsunami was highly unstable. After some period, a new balance of the coastal and river mouth condition will be achieved. The new conditions are often different from the pre tsunami condition. The recovery process may create problems. A complete closure of the river mouth occurred in anakita River, which required dredging 6). The aerial photo provides valuable information of these changes. However, the availability of the data is limited and costly. The river mouth condition may be assessed following the water level in the river relation to the tidal level. The water level in the river mouth is influenced by various other factors such as waves, winds and also the river discharge. However, the river mouth morphology at the entrance is most closely related to the tidal level. This relation has been confirmed by analytical approach 7), and numerical approach 8),9). This relation has been
successfully employed to investigate river mouth condition in Snowy River Estuary, Australia, by a combination of tidal level data with simple hydrodynamic model ). The Samegawa river mouth was investigated in this study by analyzing the aerial photo of pre-tsunami condition, immediately after the tsunami, and the most recent image. Additionally, water level analysis was conducted to assess the river mouth condition when there were no aerial photos available. 2. STUDY AREA The analysis requires the acquired images to be in the same coordinate system. The Projected Coordinate System (UTM-WGS`84 Zone 54) is used in this study. Therefore, all of the images were converted to this coordinate system before extracting the shoreline position. The study area is the Samegawa river mouth, Fukushima, Japan as shown in Fig.. The river mouth was protected with sand formation. This sand formation was completely eroded due to The Great East Japan Tsunami of 2. After the tsunami, the recovery process started and the new sand formation was formed. The longshore sediment transport moves from the South to the orth. The is located nearby. The discharge from the powerplant is separated with the river discharge by a jetty in the left bank of the river mouth, constructed in 982. Thus, sediment from this side of the river cannot pass. There were no significant damages to this jetty during the tsunami incidence. 3. RESULTS AD DISCUSSIOS () Aerial photo analysis Satellite images for the study area were collected from Google Earth for 27 April 25 (DigiGlobe, resolution.6 m), 2 March 2 (DigiGlobe, resolution.6 m), and 28 February 22 (GeoEye, resolution.5 m) which corresponds to Fig.2 (a), (b) and (d), respectively. An aerial photo was acquired on ovember 2 as shown in Fig.2 (c). a) 25/4/7 (Google Earth) b) 2/3/2 (Google Earth) c) 2// (Aerial photo) d) 22/2/28 (Google Earth) Fig. Study area. Fig.2 Samegawa river mouth aerial photos.
Y (meter) 485 4845 484 4825 483 4835 Protection X (meter) Fig.2(a), 25/4/27 年 4 月 27 日 Fig.2(b), 2/3/2 年 3 月 2 日 Fig.2(c), 2// 年月 日 Fig.2(d), 22/2/28 年 2 月 28 日 (ii) (i) Fig.3 River mouth changes based on aerial photos. The satellite images were acquired in Geographical Coordinate System (GCS-WGS`84). Thus, rectification process is not required. The acquired satellite images (Fig.2 (a), (b) and (d)) were directly projected to the desired coordinate system. However, the aerial photo (Fig.2 (c)) contains no geo-reference data. Therefore, the aerial photo was rectified to comply with the coordinate system based on several control point ). The accuracy of the satellite images and aerial photo depends on the tidal level. The digitized image accuracy ranges from 5 meters to 2 meters based on the tidal level at the corresponding date when the photos were taken. More details on the accuracy can be found in previous study 2). The digitized map of the river mouth is shown in Fig.3. It was found that the sand formation in the river mouth was completely eroded by the tsunami. The river mouth width based on the acquired image on 25/4/7 was approximately 25 meters which increased to approximately 4 meters after the tsunami incidence (2/3/2). The recovery process of the sand formation was mostly supplied from the south of the river mouth due to the existing jetty. The river mouth width was reduced to 2 meters based on the image acquired on 2//. The river mouth width returned to 25 meters in the acquired image on 22/2/28. However, the newly formed sand barrier retreated approximately 2 meters from the pre-tsunami formation (Fig.3 (i) and (ii)). (2) Water level analysis The water level measurement and the discharge data in the river are available at the upstream of the river mouth as shown in Fig.. The tidal level data is available at the Onahama port. The data were converted to the same datum based on the elevation of the station. Measurements of the water level in the Samegawa River, tidal level, and discharge for the period of 2/6/-6/8 and 22/3/2-3/6 are given in Fig.4 (a) and (b), respectively. Interesting phenomenon can be seen in these figures. It is clear that the tidal level fluctuation is reflected to the water level measurement in the river. Their relation will depend on the river mouth condition and the river discharge. After the tsunami (Fig.4 (a)), the river mouth was wide open, hence the tidal exchange into the river was not reduced and delayed. However, as the recovery process started and the river mouth was gradually constricted, the effect of the tidal level in the river was reduced and there can be time delayed between the water level in the river and the tidal level at the sea (Fig.4 (b) ii). Fig.5 shows the hourly relation between the water level in the river and the tidal level. It was found that both data shows good linear relation during low and high tide, after the tsunami up to the end of 2. However, this relation changes as the river mouth gradually closed causing the water level in the river (η R ) can not closely follow the tidal level (η O ) as shown in Fig.5 (i)). In this study, the above-mentioned relation is investigated. water level (m) water level (m) 6 5 4 3 2 - -2 4 2/6/ 2/6/2 2/6/3 2/6/4 2/6/5 2/6/6 2/6/7 2/6/8 6 5 4 3 2 - (i) water level threshold (a) 2/6/-6/8 (ii) tidal level Samegawa River River Discharge -2 4 22/3/2 22/3/3 22/3/4 22/3/5 22/3/6 (b) 22/3/2-3/6 tidal level Samegawa River River Discharge (i) water level threshold Fig.4 The Samegawa water level, tidal level, and discharge 5 5 2 25 3 35 5 5 2 25 3 35 Q (m 3 /s) Q (m 3 /s)
i -.6 -.4 -.2.2.4.6.8 η O (m) 2/6/24-6/3 (average).4.2.8.6.4.2 Fig.5 Water level relation to tidal level ηr (m) 6//2 //2 2//2 //22 2//22 3//22 3/5/22 4//22 4/5/22 Table. a and R verification (anakita River) 6) Date a R.84.98 2/7/.82.97 2/8/9.23.44 2/8/- 9/2 (average).8.2 Condition (aerial photo) River mouth was wide open after the tsunami Gradual closure of the river mouth due to recovery of the sand formation Full blockage of the river mouth 2/9/22 Flood 2/9/23-9/3 River mouth was (average).42.8 re-opened after the flood The correlation coefficient (R) between the water level at the river mouth and the tidal level at sea was calculated, considering the modification of water level variation in the river mouth with the decrease of the cross-section. Additionally, a gradient coefficient (a) is calculated which is defined as the gradient of the linear regression of the water level data following the relation bellow. η R =a η O + b () where η R is the water level in the Samegawa river mouth, η O is the tidal level in the sea, and b is the constant. It should be noted that during the period of high river discharge, the accuracy may be reduced since the high discharge causes an increase of the water level in the river that may not be reflected in the tidal level variation. The method has been verified by analysing the river mouth changes in the anakita river. The anakita river mouth has similar condition with the Samegawa river. Both of their river mouths were located in sandy area with sand formation in front of the river mouth. In addition, more image data were available for the anakita river to confirm the relation between the proposed parameters of the correlation coefficient (R) and the linear gradient (a) to the condition of the river mouth. It has been reported that the proposed parameters correlate well with the anakita river mouth condition 6). The recovery process after the tsunami of 2 caused the anakita river mouth blockage that occurs gradually from the beginning of July to August, 2. A complete river mouth blockage that occurred caused almost no tidal intrusion hence the low value of R and almost no value of a. The anakita river mouth was re-opened again after a flood which caused the re-occurrence of tidal exchange in the river hence the increase of a and R value. The verification resume is given in Table. In this study, analysis was conducted to the water level data during the period of 2/6/ to 22/6/7. However, the tidal level port was not recorded during the period of 2/6/26 to 22/3/2. Therefore, the water level data during this period were replaced with the astronomical tide provided by the port. Additionally, the water level measurement device in this river may not accurately record the very low water level due to the nature of the measurement (threshold in Fig.4 (a) and (b), (i)). It was found that the threshold value may be different after a period of high discharge. The slight fluctuation during this low water level does not reflect the actual value. Thus, a % fluctuation in this situation is not considered in the analysis. The linear gradient (a) value in Fig.6 shows significant differences between the analyzed periods. The a values in the period of (), (2) and (5) are approximately.9. During these periods, the river mouth was opened. Therefore, tidal exchange may occur including in low tide which is confirmed by the high value of a. However, the a value for the period (3) is.75 and the period (4) is.43. During these period, river mouth was constricted hence tidal exchange was restrained. These correspond to the recovery process of sand formation as it was found in the aerial photo. (4) a =.43 (3) a =.75 (2) a =.88 () a =.88 (5) a =.92 -.8 -.4.4.8 η O (m) () 22, June 8-4 (2) 2, August 8-4 (3) 22, February -7 (4) 22, April -7 (5) 22, May 4-2 Fig.6 Linear gradient (a) coefficient..6.2.8.4 ηr (m)
Fig.7 shows a detail daily analysis of each parameter. Here, Q is the average of the hourly discharge in day (Fig.7 (a)). The daily value of R and a are given in Fig.7 (b) and (c), respectively. Overall, both parameters show similar behavior, which relate to the river mouth condition and discharge. The river mouth was widely open after the tsunami. The wide opening of the river mouth causes almost no lag as well as no reduction in the tidal response in the river, even during low tide. Hence, a high value of R and a can be observed. This agrees well with the image analysis for the acquired image on 2/3/2 (Fig.2 (b)). Their value decreased gradually, starting from September 2 until May 22, with a rapid decrease started in January 22. This agrees well with the aerial photo (Fig.2 (c) and (d)). However, the minimum value of these parameters during the analyzed period for the Samegawa River was still high as compared to the value for the anakita River. Thus, there was no full river mouth blockage although the recovery process caused the narrowing of the river mouth as shown in Fig.2 (c) and Fig.2 (d). The tidal exchange still occurred in the Samegawa River but there can be time lag, reduction, or tidal restrained. The flood in 22/5/3 re-opened the river mouth. The river mouth was no longer constricted. There was no image to confirm this incidence, however as it has been explained earlier, a wide opening of the river mouth will cause no significant time lag nor amplitude reduction in the tidal exchange even during low tide, causing the high value of R and a. 35 3 Q Daily Average (m 3 /s) 25 2 5 5 2/6/ 2/6/29 2/7/27 2/8/24 2/9/2 2//9 2//6 2/2/4 22// 22/2/8 22/3/7 22/4/4 22/5/2 22/5/3 22/6/27 Date (a) Discharge in Samegawa river.8.6 R.4.2 2/6/ 2/6/29 2/7/27 2/8/24 2/9/2 2//9 2//6 2/2/4 Date 22// 22/2/8 22/3/7 22/4/4 22/5/2 22/5/3 22/6/27 (b) Correlation coefficient (R).8.6 a.4.2 2/6/ 2/6/29 2/7/27 2/8/24 2/9/2 2//9 2//6 2/2/4 22// 22/2/8 22/3/7 22/4/4 22/5/2 22/5/3 22/6/27 Date (c) Linear gradient coefficient (a) Flood Fig.6() Fig.6(2) Fig.2(c) Fig.6(3) Fig. 2(d) Fig. 6(4) Fig. 6(5) Fig.7 Water level analysis.
4. COCLUSIOS The morphology changes in the Samegawa River mouth was investigated in this study using aerial photo and water level measurement. The aerial photo analysis shows the opening of the river mouth after the tsunami. The river mouth width was greatly increased from 25 meters to 4 meters. The gradual recovery process of the sand formation in front of the river mouth was also detected. The river mouth width was reduced to 2 meters in ovember, 22 and then to 25 meters in February 28, 22. In addition, the new sand formation was formed approximately 2 meters upstream of the previous formation. The tidal exchange behaviors were investigated to assess the river mouth condition where there were no image data available. A wide opening of the river mouth causes no time lag and no reduction of the tidal exchange amplitude, regardless of low tide or high tide. However, the constriction/blockage of the river mouth may cause a time lag between the river water level and the tidal level, as well as tidal exchange restraint during low tide. These phenomenon were observed in the Samegawa river and assessed using the correlation coefficient (R) and the linear gradient coefficient (a). The assessed parameters correlates well with the satellite image analysis showing a gradual decrease of the parameters from September 2 to May 22 that corresponds to the river mouth constriction due the recovery process of the sand formation. The flood in 22/5/3 re-opened the river mouth. This process caused less obstruction in the tidal exchange, which was confirmed with the increase of a and R value. This study has shown the assessment of river mouth condition based on aerial photo and water level analysis. In addition, the applicability of water level analysis has helped to assess the river mouth condition when there are no aerial photos data. Therefore, this method will be very useful for assessing the morphology changes in the river mouth. ACKOWLEDGMET: The authors wish to express grateful thanks to the Miyagi Prefectural Government and the Ministry of Land, Infrastructure and Transport for providing us the precious water level data. This research could not be conducted without financial supports from JST/JICA, SATREPS (Science and Technology Research Partnership for Sustainable Development), the Grant-in-Aid for Scientific Research from JSPS (o. 23623, o. 223693, o. 23367), Grant-in-Aid for Scientific Research from the River Environmental Fund (REF) in charge of the Foundation of River and Watershed Environmental Management (FOREM), as well as Assistance for Technological Development, Tohoku Construction Association. The first author is JSPS Postdoctoral fellow no. P367. The authors would like to gratefully appreciate their supports. REFERECES ) Mori,., Takahashi, T. & The 2 Tohoku Earthquake Tsunami Joint Survey Group: ationwide post event survey and analysis of the 2 tohoku earthquake tsunami, Coastal Engineering Journal, Vol. 54, o., 22. 2) Adityawan, M., B., Roh, M., Tanaka, H., Mano, A., Udo, K.: Investigation of tsunami propagation characteristics in river and on land induced by the great east japan earthquake 2, Journal of Earthquake and Tsunami, 22. (in press) 3) Tanaka, H., guyen, X. T., Umeda, M., Hirao, R., Pradjoko, E., Mano, A. and Udo, K.: Coastal and estuarine morphology changes induced the 2 Great East Japan Earthquake Tsunami, Coastal Engineering Journal, Vol.54, o., 2. 4) Udo, K., Sugawara, D., Tanaka, H., Imai, K., Mano, A.: Impact of the 2 Tohoku Earthquake and Tsunami on beach morphology along the northern Sendai Coast, Coastal Engineering Journal, Vol. 54, o., 22. 5) Adityawan, M., B., Roh, M., Tanaka, H., Farid, M.: The effect of river mouth morphological features on tsunami intrusion, Proceedings of International Conference on Disaster Management, 22. (CD-ROM) 6) Tanaka, H., Adityawan, M., B., Mano, A.: River mouth morphology changes at the nanakita river mouth after the great east japan tsunami, Annual Journal of Coastal Engineering, B2-68, 22. (in Japanese) 7) Keulegan, G.H.: Tidal Flow in Entrances: Water Level Fluctuations Of Basins in Communication With The Seas. Committee on Tidal Hydraulics Technical Bulletin, o. 4. 967. 8) Shuttleworth, B., Woidt, A., Paparella, T., Herbig, S., Walker, D.: The dynamic behaviour of a river-dominated tidal inlet, River Murray, Australia, Estuarine, Coastal and Shelf Science, Vol. 64, pp. 645-657, 25. 9) Oliveira, A., Fortunato, A.B., Rego, J.R.L.: Effect of morphological changes on the hydrodynamics and flushing properties of the O bidos Lagoon (Portugal). Continental Shelf Research, Vol. 26, pp. 97-942, 26. ) McLean, E.J., Hinwood, J.B.: Modelling entrance resistance in estuaries. Proc. 27th International Conference on Coastal Engineering, pp.3446-3457, 2. ) Pradjoko, E. and Tanaka, H. (2): Aerial photograph of Sendai Coast for shoreline behavior analysis, Proc. 32nd Int. Conf. Coastal Engineering. 2) Adityawan, M.B., Tanaka, H., agabayashi, H.: Investigation of shoreline change caused by the Great East Japan Tsunami 2 around Samegawa River mouth. Proceedings of International Sessions of Coastal Engineering Conference, Vol. 3, pp. 3-35, 22. (Received September 3, 22)