MES-190301.pdf4.46MB

1. Introduction

 Jakarta is the capital city of Indonesia which is located at the northern coast of the West Java Island. The area is relatively flat, with the topographical slopes ranging between 0° and 2° in the northern and central parts, and between 0° and 5° in the southern part. The southernmost area of Jakarta has an altitude of about 50 m above mean sea level(Abidin et al., 2009).

 Jakarta Bay has been suffering from chronic problems of coastal flooding, poor infrastructure, solid wastes, and pollution among others(Pasang et al., 2007; Abidin et al., 2009). These problems have been the main issues to be addressed by the government authorities. Especially, coastal flooding in Jakarta is a complex and serious problem because flooding is happening through compounding effects of sea level rise, land subsidence, waste deposition to water canals and so on. In Jakarta, land subsidence has been heavily blamed for the primary cause of flooding.

 Over the period of 1982-1997, land subsidence ranging from 20 to 200 cm is evident in several places in Jakarta. In general the land subsidence exhibits spatial and temporal variations, with the rates of about 1 to 15 cm/year(Abidin et al., 2009). A few locations exhibit subsidence rates up to 20 - 25 cm/year. It was found that the spatial and temporal variations of land subsidence depend on the corresponding variations of groundwater extraction, coupled with the characteristics of sedimentary layers and building loads above it.

 Land subsidence in Jakarta can be caused by four factors, namely: excessive groundwater extraction, load of buildings and constructions, natural consolidation of alluvium soil, and tectonic activities. Up to now, there is no information yet about the contribution of each factor on the subsidence at each location and their spatial (contribution) variation. In case of Jakarta, tectonic activities seem to be the least dominant factor, while excessive groundwater extraction is considered to be one of dominant factor. The first three factors will have a close relation with urban development activities in Jakarta and its surrounding areas(Abidin et al., 2009). In case of Jakarta, which is actually prone toward flooding, subsidence phenomena has to be fully understood for flood management system. During the period between 1993 and 2007, at least there were four big floods in Jakarta on January 9 - 10, 1993; February, 1996; January 26 - February 1, 2002; February 4 - 14, 2007.

 Typically, land subsidence studies have been carried out using GPS study which utilized multiple data around the study areas and presented as Figure 1 which shows contours of land subsidence in 2D format over the period of 1982-1991 (left) and 1991-1997(right)(Abidin et al., 2001).

Fig. 1. Land subsistence in Jakarta(Source: Abidin et al., 2001).

 The problem associated with such presentation of data is that the figure would not be easily understood by policy makers and laymen or seldom create impacts out of the scientific research. Policy makers seldom feel this would be serious problem as they have less sensitivity on the number it represents. To overcome this weakness of scientific presentation, the Terrestrial LIDAR (LIght Detection And Ranging) survey has been employed to improve visual effect of scientific information for better understanding of the actual situation of the subsidence phenomena in Jakarta Bay. Terrestrial LIDAR data is actually a photographic image with coordination data embedded in the data points.

 The purposes of this research are to examine the use of Terrestrial LIDAR in identifying the characteristics of land subsidence in Jakarta Bay and its application for identifying flood vulnerable areas and to generate scientific information (in 3D format) which will be easily understandable to decision makers. The research team conducted Terrestrial LIDAR surveys in a demonstration site, Pantai Mutiara where significant subsidence has been reported.

2. Study Site

 Figure 2 shows the study site of this research, Pantai Mutiara. The site is a newly established luxurious residential area which has several high-rise condominiums. Almost all the roads are paved with asphalt and concrete equipped with side drainage. Most of  houses have mooring facilities for yachts. Coastal line is also constructed with low dike. Figure 3 shows the photo of west main road of Pantai Mutiara (shot taken from spot A of Figure 2) which clearly shows that the sea level is almost even with the residential area.

Fig. 2. The satellite image of the study site (white box)(satellite image : maps.google.com).

Fig. 3. A photograph of west main road of the Pantai Mutiara (A area in Fig. 2).

 According to Lee et al.(2003), Pantai Mutiara area has been reclaimed in 3 stages as shown in Figure 4(a). The first stage (A area in Figure 4) was completed from 1986 to 1988. Second stage (B area) was in 1994 and the third stage (C area) was by 2007. The Terrestrial LIDAR surveys have been conducted in area B in Figure 4. The Figure 4(b) shows area map with elevation information included after the completion of the stage 1 (A area) reclamation. The elevation of one point in area A is 1.84 m and one point in area B is 5.78 m.

Fig. 4. Reclamation stages of Pantai Mutiara (a); and elevation map after the completion of stage 1 (b).

3. Methods

 LIDAR is an abbreviation of ‘LIght Detection And Ranging’ which uses a LASER (Light Amplification by Stimulated Emission of Radiation) beam. LIDAR obtains data through scanning of a site. One scan usually generates more than several million to tens of million points. These points are called as ‘point cloud’. Data points usually comprise of coordination information (x, y, z), response intensity, and RGB color value of each point. The coordination information of each data point determines the relative position of LIDAR which will be normalized for geographical coordination such as UTM (Universal Transverse Mercator) through correction with a DGPS (Differencial Global Positioning System) or a Total Station data at the reference points(Park, 2008). The LIDAR technology can secure data with millimeter precision while currently available Digital Elevation Model (DEM) data in Jakarta Bay is in centimeter precision scale. In this research, the LIDAR equipment used was a Terrestrial LIDAR which was operated on a tripod which was manufactured by Optech (model: ILRIS 36D).

 The Terrestrial LIDAR survey was conducted by the three step sequence: a survey design, survey and data processing. Design stage includes a set-up of reference (control) points, ranging of scanning area, and identification of area characteristics through pre-scan. After the scanning, data processing was followed which included data alignment, geometric correction, removal of errors and editing of Terrestrial LIDAR data.

 The scanning of Terrestrial LIDAR was conducted at the rooftop of a 27-story condominium (inset of Figure 5) so as to secure wide area on October 3 and 4, 2011. Based on the data obtained from pre-scan, possible scanning range and density of data were identified. Also to compensate the loss areas due to shades, additional 38 scanning points were selected which were scanned at the ground level on October 5, 2011 (Figure 5). In figure 5, SP denotes the scanning point where Terrestrial LIDAR scanner was installed. Total of four scans starting from west to east were conducted (A_01, A_02, A_03 and A_04) in order to cover the entire area. Each scanning took around 2 hours.

Fig. 5. Scanning of the area. SP denotes the scanning point which is a roof-top of 27-story building (inset). P denotes the reference point. Numbered arrows indicates the additional scanning at the ground.

 Adjusting data sets scanned various position into one standard coordination system is called ‘alignment’. The four data set of LIDAR scanned on rooftop of condominium have been aligned using coordinates of the control points (P_1, P_2, P_3, P_4 and P_5 in Figure 5).

 Rooftop scanned data and ground scanned data have been aligned using a statistical method (best-fit). The best-fit alignment uses distance minimizing method and processing step is as follows; first, calculating all distance of each data set's points. second, changing pose of 1 data set. third, re-calculating all distance of points. forth, comparing each sum of distances of before and after changing pose. fifth, choosing pose with smaller sum of distances. sixth, repeating second to fifth step until sum of distance is smaller than user defined tolerance error (Figure 6).

Fig. 6. Example of statistical alignment method.

 The 4 high-density rooftop scanned data and 38 ground scanned data were then verified by comparing with 60 validation points measured by total station in non-prism mode. Figure 7 shows the distribution of deviation of 3D coordination compared with the validation points. Red color points indicate deviation of more than 5 cm. The result of verification showed that almost all the validation points have less than 2cm deviation whereas only 2 points have more than 5 cm deviation. The root mean square error (RMSE) is 1.91 cm. Instrumental error of LIDAR is about 0.7 cm and Total-station is 1.5 cm (non-prism mode). Therefore, we can conclude that there is a negligible error in the alignment.

Fig. 7. Distribution of deviations from validation points.

 For correcting vertical position by MSL (Mean Sea level), GPS data measured during 6 hours and the Tanjung Priok tide data used (Table 1). The measured GPS raw data was processed using the reference data of the BAKO GPS station that is the Indonesian zero order geodetic point.

Table 1. Non-harmonic constants of the Tanjung Priok tide station

4. Results and Discussion

4.1 Distribution of Elevation

 The Figure 8(a) and (b) are showing actual photograph of the site (a) and its corresponding Terrestrial LIDAR data image (b). Although Figure 8(b) looks like a black and white picture of Figure 8(a), it is actually collection of data points which have their own coordination information. From the Terrestrial LIDAR data, we can cull out only necessary data sets to meet the purpose of the study. The selected data sets can be easily tagged with color system as shown in Figure 9.

Fig. 8. Picture of the survey area taken in a digital camera (a) and the corresponding Terrestrial LIDAR data image (b).

Fig. 9. 2-dimensional Terrestrial LIDAR data of survey site: (a) all the structures; and (b) the roads in the dotted area in (a).

 Figure 9 shows 2-dimensional coordination (x and y) with elevation in color where the blue color represents areas below mean sea-level and other colors represent above mean sea-level. South-east corner of the dotted box in Figure 9(a) has areas below sea level and other areas are in 1 or 2 m above sea level. Figure 9(b) highlighted the elevation of the roads in the survey site. Since most of the real estates in Pantai Mutiara are owned by individuals, the owners of the land are trying to heighten their lands due to land subsidence. However, in the case of roads, the owners of the land cannot modify the roads since they are publicly owned. Therefore, we may safely assume that roads may represent the original state of the area and it would be meaningful to look into only roads for the analysis of land subsidence.

 The image of roads in the survey area was formatted with colors which represent their relative elevation compared to mean sea-level as shown in Figure 9(b) with which 5 roads (A to E) were examined. Table 2 shows highest, mean, lowest elevation and standard deviation of roads in each section (Figure 9(b)). Mean elevation of all roads is 0.24 m that is lower than the approximate highest high water 0.51 m.

Table 2. Elevation of roads in each section

 Figure 10 shows cross-sectional elevation of those 5 roads in Figure 9(b) in 2-D format (a) and 3-D format (b). In the case of road A-A’, the height difference was within 10 cm range along 380 m-length road whereas in the case of road E-E’ the height difference was more than 60 cm. To our dismay, it was found that significant portions of the roads D-D’ and E-E’ were below sea-level as shown in blue color in Figure 9(b). Therefore, we can see that land subsidence is seriously happening in D’ and E’ areas where lands are below mean sea-level. The 3 dimensional elevation of the roads as shown in Figure 10(b) clearly demonstrates that the spot subsidence rate vary significantly from area to area.

Fig. 10. (a) 2-D elevation of the 5 roads; and (b) 3-D elevation of roads with color representing the elevation with respect to mean sea-level.

4.2 Prediction of Flooding Areas using Sea Level Rising and Land Subsidence

 IPCC reported that the sea level will rise up to 59 cm at the end of this century(Solomon et al. 2007). For Indonesia, it was reported that approx. 2 - 3 mm/yr up to 7 mm/yr might be possible. Since Jakarta is showing higher rate of land subsidence than sea level rise, it is important to continuously monitor the rate of subsidence in order to devise proper policy measures. Therefore, it is necessary to consider both sea level rise and subsidence to properly predict possible inundation and flooding in the future.

 In this study, the subsidence rate of about 6 cm/yr was assumed according to various literature(Abidin et al., 2009). Sea level rise is assumed about 2.1 mm/yr which was calculated from Jakarta tide station data from 1992 to 2004 (UHSLC; University of Hawaii Sea Level Center). Using these rates, scenarios for 1 year, 5 years and 10 years have been developed. For each scenario, flooding height was set at mean sea level and highest tide height (Table 3). In Figure 9, inundated roads are presented in each scenarios: (a) 1 year; (b) 5 years; and (c) 10 years.

Table 3. Flooding height considering land subsidence and sea-level rise

 As we can see in Figure 11, almost all roads will be inundated with sea water due to sea level rise and land subsidence of the area within 10 years. This result is not quite surprising because several areas in Pantai Mutiara is already below mean sea level as of now.

Fig. 11. Predicted flooding roads due to land subsidence and sea-level rise (red: flooding area by mean sea level, yellow: by highest high water) : (a) 1 year; (b) 5 years; and (c) 10 years.

4.3 Analysis of Dike Height

 Through the survey and analysis, we were able to find that Pantai Mutiara area is highly vulnerable of coastal flooding that is caused mainly by high subsidence rate rather than sea level rising. The last stand which keeps the area from flooding and inundation are the dikes built around the area. Although dikes are still functioning, the height of dikes are getting lower and uneven due to subsidence and partial reconstruction. We can conclude that flooding vulnerability of this area is depending of the height of dikes.

 Figure 12 shows the height of dikes 2 dimensional color format (a) and in 3 dimensional view (b) of the rectangular area in (a). In Figure 10(a), red color indicates areas where dike is completely submerged under water and blue color indicates the areas where dike is maintaining its height. In this figure, we can see that there are five places where dikes are completely submerged and shoreline is intruding backyard of houses. Those five areas where dikes are under sea-level, reconstruction will be required immediately in order to prevent any disaster in near future.

Fig. 12. Distribution of height of dike (unit: m) in 2D format (a); and 3D format (b).

5. Conclusions

 In this study, 3D precision Terrestrial LIDAR survey has been conducted to review land subsidence problem at Pantai Mutiara, in Jakarta Bay. With the study we can identify that the survey site has mean elevation of 0.24 m with the highest elevation of 0.93 m and lowest –0.35 m. Considering that highest tidal height is 0.51 m, many areas of the survey site are lying below the highest tidal height. Therefore, Pantai Mutiara area is seriously vulnerable to inundation and flooding. Thus, it is important to continuously monitor subsidence as subsidence is continuously occurring as of now.

 Also, the study could demonstrated that Pantai Mutiara area is showing various subsidence rates depending on sites although the site is relatively narrow and small. There is elevation differences of almost 1m within the survey site. We found that west section is higher than east section as the roads are subsiding with significantly different rates.

 In our study, the flooding risk was identified incorporating the compounding effects of sea-level rise and land subsistence. We found that dikes may not function properly within 10 years as flooding and inundation will be severe due to poor drainage of sea water. Also, since height of dikes are different from area to area, there is a possibility that inundation may happen at the low dike site.

 In this study, one time Terrestrial LIDAR survey was able to provide key information including topography, dike height distribution, and future flood risk of the survey area. Especially, as the 3D precision topography effectively conveys important messages relating to vulnerability of the site, policy makers and layman can easily understand the situation of the site. In this, we found that Terrestrial LIDAR technique is an efficacious method in identifying coastal vulnerable areas due to land subsidence and sea-level rise.

Acknowledgements

 This research was supported by the KIOST project “Study on Establishment of Integrated Coastal Management Program in Jakarta Bay Area, Indonesia” (PG47920) funded by the Organizing Committee of the Yeosu Expo 2012. The administration was supported by KOICA.