Beach Changes Triggered by Ground Subsidence and Waves Generated by Easterly Trade Wind near Pekalongan on North Shore of Java Island

Abstract

The shoreline recession caused by ground subsidence due to the excessive pumping of underground water and that caused by wind waves incident from the easterly direction under the tropical monsoon weather was investigated, taking the Pekalongan area on Java Island as a study area. The study area was divided into nine subareas on both sides of the Comal River delta, and the shoreline changes in each subarea were investigated using satellite images. Field observation was also carried out on August 10, 2022, in the Pekalongan area, and the height of ground subsidence was measured together with the observation of the inundated area using a UAV. It was found that, in this area, the height of ground subsidence reached approximately 1.6 m, and the sinking area has been rapidly expanding. In this area, predominant waves are constantly incident from the easterly direction under the tropical trade wind condition, and in some cases, waves are obliquely incident at a significantly large angle relative to the direction normal to the shoreline. Along the east coast of the Comal River mouth, the mean wave incidence angle was as large as 22° clockwise, whereas on the west side, the angle increased up to 42°, causing marked shoreline changes due to high-angle wave instability.

1. Introduction

A substantial proportion of the world’s sandy coastline is already eroding, and such a situation may be exacerbated by climate change. Vousdoukas et al. [1] showed that these ambient trends in shoreline dynamics, combined with coastal recession driven by sea level rise, could result in the near extinction of almost half of the world’s sandy beaches by the end of the century. Mentaschi et al. [2] gave a global and consistent evaluation of coastal morphodynamics between 1980 and 2015 based on satellite observations and concluded that the observed global trend in coastal erosion could be enhanced by future sea level rise along with the discussion on the impact due to anthropogenic factors. Thus, if a sea level rise develops, there will be an acceleration of existing beach erosion [3]. The changes predicted globally are already in evidence on sections of the world’s coastline where land subsidence has produced a relative sea-level rise. Thus, the ground subsidence has the same effect as the sea level rise, implying that the study on ground subsidence is important to understand the impact of sea level rise to the coastal lowland. This could be done by a fact-finding investigation on real coasts.

Here, we aim to study the real example occurring between Pekalongan and Pemalang in Central Java, Indonesia, where severe ground subsidence has already occurred not only by the excessive pumping of underground water but also wave-induced sand transport under the constant tropical trade wind; topographic changes due to ground subsidence and wave action have simultaneously occurred. When a non-uniform ground subsidence occurs along the coastline, the longshore inclination of the shoreline changes, resulting in the occurrence of longshore sand transport, even though waves are incident normal to the previous shoreline. In this manner, the study on the combined effect of the ground subsidence and longshore sand transport to the topographic changes along a specific coast is no less important compared with the study on the global effect by sea level rise.

The coastline in the present study area between Pekalongan and Pemalang runs approximately in the east-west direction, and the shoreline protrudes into the Java Sea at the Comal River delta (Figure 1). Along this river delta coastline, waves may be incident at a large angle relative to the direction normal to the shoreline because of the protrusion of the shoreline, causing significant beach changes.

Figure 1.

Satellite image of the study area between Pemarang and Pekalongan, taken on 29 September 2021.

As regard to the ground subsidence in this area, Chaussard et al. [4] estimated the rate of ground subsidence in Pekalongan to be 3–10 cm/yr. in 2013. Then, Andreas et al. [5] collected the evidence for subsidence based on repeated leveling measurements, GPS surveys, InSAR (Interferometric Synthetic Aperture Radar) measurements, extensometer, etc., and showed that the average rate of subsidence was as large as 0.01–0.1 m/yr. in 2018. They also pointed out that although dykes have been built against tidal inundation around the coasts on north Java coast, the dike subsiding through times results in the failure in their capability against ground subsidence and sea level rise, because the land subsidence and the sea level rise continue through times. Sarah and Soebowo [6] presented a brief review on land subsidence along the north coast on Java Island, and they showed the main features of the land subsidence including the subsurface geology, land subsidence monitoring, and analysis of possible causes and mechanism. Their time series analysis of InSAR images revealed that subsidence in Pekalongan occurred at a rate of 4.8–10.8 cm/year, suspected to be caused by groundwater withdrawal for agriculture.

Moreover, Andreas et al. [7] showed again in 2020 remote sensing capabilities on land subsidence and coastal water hazard using the GNSS (Global Navigation Satellite System) and InSAR data and concluded that, in the Pekalongan area, the ground subsidence started since late 1970s and the entire ground subsidence reached 3.7 m until 2015.

Sidiq et al. [8] also investigated the ground subsidence of the entire Java Island using SAR satellite images and argued that the height of ground subsidence in Pekalongan reached 22 cm between 2017 and mid-2020. Thus, most of the studies focus on the rate of ground subsidence in the inland area of Pekalongan, and studies on the interaction between cumulative ground subsidence and beach changes caused by the imbalance in longshore sand transport due to waves are rare.

Similar ground subsidence triggered by the excessive pumping of underground water has occurred in the southern part of Kujukuri coastal plain in Chiba Prefecture, Japan, and the maximum ground subsidence reached 90 cm between 1968 and 2012, causing the shoreline recession of 18 m when assuming the foreshore slope of 1/20 [9]. When the effect of sea level rise associated with the greenhouse effect occurs with the ground subsidence, enormous damage may occur. Therefore, in this study, beach changes were investigated, taking the suburbs of Pekalongan as an example, where coastal lowland vulnerable for coastal damage develops along the coastline, and both the ground subsidence and the shoreline recession caused by the imbalance in longshore sand transport due to waves simultaneously have occurred. Through the study, we investigate not only the expansion of the inundation area associated with ground subsidence but also the reversal in the direction of longshore sand transport caused by local ground subsidence. Moreover, along the study area, the angle of the predominant waves takes a value greater than 45°, possibly causing high-angle wave instability. We also discuss a measure to prevent the devastation of the coastline along such a coast.

First, nine subareas were adopted alongshore between Pekalongan and Pemarang located 20 km west of the Comal River delta (Figure 1), and the shoreline changes in each subarea were investigated using Google Earth, together with field observation on 10 August 2022 near Pekalongan. In Figure 1, the lowland shown in black extends in the vicinity of the shoreline in subareas 1–4 west of Pekalongan where significant ground subsidence occurred and particularly large ground subsidence occurred in subareas 2 and 4.

The results of the study in each subarea are described in Chapter 3. Then, in Chapter 4, the longshore distribution of the angle relative to the direction normal to the shoreline is shown, which closely relates to the longshore sand transport on each coast. Then, discussions were made in Chapter 5.

Regarding the wave characteristics in this area, wave height was measured by Sakuno et al. [10] using the satellite sea level altimeter. They showed that the significant wave height with a high probability of occurrence in the Java Sea ranges between 0.5 and 1.0 m, relatively low with a wave period shorter than 4 s. Furthermore, the significant wave height H₉₅ with the non-exceedance probability of 95% was 1.48 m.

2. Method of study

The Comal River delta protrudes into the Java Sea in the central part of the study area (Figure 1). In this area, subareas 1–4 were set westward from Pekalongan east of the Comal River delta, whereas subareas 6–9 were set alongshore until Pemalang west of the river delta along with subarea 5 at the center of the Komal River delta. The shoreline of the Comal River delta has an asymmetric shape with respect to the centerline of the delta, as shown in Figure 1. A straight line is drawn in the east–west direction at a location 4.2 km south of the river mouth, and the direction normal to the shoreline at the intersections between this line and the shoreline is calculated; the angle of the direction normal to the shoreline with respect to the north becomes 50° (N50°E) on the east side and 41° (N41°W) on the west side of the delta. Thus, the angle of the direction normal to the shoreline relative to the north on the east side (50°) is larger than that of 41° on the west side of the river delta. This asymmetry of the river delta is assumed to be due to the asymmetric effect of waves incident from the eastern side of the river delta under the constant action of easterly wind. Because when waves are incident from the direction of the center line dividing a symmetric river delta into two, a symmetric river delta must be formed due to the symmetric action of waves.

In this study, the suburbs of Pekalongan (subarea 2) where severe ground subsidence occurred, the Sikebu River mouth (subarea 4), the Comal River delta (subarea 5), and subarea 6 west of the river delta, as shown in Figure 1, were selected out of nine subareas, and the shoreline changes in each subarea were investigated. The distribution of wave incidence angle relative to the direction normal to the shoreline was measured in these nine subareas because oblique wave incidence is vital to cause the longshore sand transport.

3. Results of the study in each subarea

3.1 Subarea 1 (suburbs of Pekalongan)

Figure 2 shows a satellite image of the area immediately west of Pekalongan taken on 29 September 2021. The residential area of Pekalongan extends along the left bank of the Banger River. Severe ground subsidence occurred in this area, and the urban area is surrounded by the submerged area shown in black. In addition, jetties of approximately 300 m length have been constructed on both sides of the Banger River, and the shoreline advances most on the east side of the jetty, implying that westward longshore sand transport prevails in this area.

Figure 2.

Satellite image of the area near Pekalongan and detailed study area.

On the other hand, the shoreline was assumed to retreat at the west end of the seawall because the shoreline in the urban area west of the river mouth is fixed by the seawall. Therefore, when adopting a rectangular area in Figure 2 west of the urban area, satellite images taken five times between March 2003 and September 2021 were compared (Figure 3).

Figure 3.

Satellite images of subarea 1 at Pekalongan.

In March 2003, no clear evidence of erosion or effects of ground subsidence were observed, and the shoreline extended straight in the east-west direction, and a sandy beach extended alongshore with neatly divided gardens and fish ponds in the hinterland (Figure 3(a)). However, by September 2009, groins and a seawall were constructed, as shown in Figure 3(b), east of the crematory indicated by a circle in Figure 3(a) as a measure against erosion. Then, because of the ground subsidence and the blockage of westward longshore sand transport, the shoreline was retreated by 54 m at the west end of the seawall by August 2013 (Figure 3(c)). Here, in Figure 3(c), the propagation of swell waves from the northeast can be clearly identified; therefore, the direction normal to the wave crests immediately offshore of the wave breaker zone near the crematory was determined to be N36°E. Because the direction normal to the shoreline in this vicinity is N14°E, waves are obliquely incident clockwise at an angle of 22° relative to the direction normal to the mean shoreline, resulting in the predominance of northwestward longshore sand transport and erosion at the west end of the seawall. Moreover, ground subsidence was intensified between 2009 and 2013, and footpaths of the fish pond became invisible.

Then, a straight seawall was constructed at a location 30 m offshore of the shoreline west of the crematory as a measure against erosion until April 2014 and connected to the seawall constructed earlier (Figure 3(d)). However, these seawalls lost their function because of ground subsidence, and they were destroyed by September 2021 (Figure 3(e)). Simultaneously, severe ground subsidence occurred, and farmlands and fish ponds that extended in the coastal lowland in 2003 were submerged under the sea surface and the facilities of the crematory remained under the sea.

As mentioned above, marked topographic changes occurred between March 2003 and September 2021 in subarea 1. The situation of the ground subsidence was confirmed on August 10, 2022, by the field observation. Figure 3(e) shows the location (Sts. 1–9) where photographs were taken. First, at St. 1 shown in Figure 3(e), the difference in elevation between the sea surface at the subsided land on the right (west) side of the seawall and the surface of the coastal road on the left was measured using a measuring stick (Figure 4(a)). As a result, the elevation of the road surface was found to be 0.2 m lower than that of the sea surface at that time. As the tide level at the measurement time was the mean sea level (MSL) + 0.10 m, the ground elevation of the road was MSL − 0.10 m.

Figure 4.

Photographs showing ground subsidence taken at Sts.1–9 in Pekalongan.

Similarly, Figure 4(b) shows the measurement of the ground elevation on the east side of the road, and the ground elevation was found to be 0.8 m lower than the elevation of the road. Although the elevation of the road has been raised after the successive ground subsidence, the ground elevation of the residential area has not been changed; these indicate that the ground elevation in this vicinity is MSL − 0.90 m. As the ground elevation before the ground subsidence was not known, the accurate ground subsidence could not be determined. However, when the original ground elevation of the coastal lowland was assumed to be equal to the berm height of MSL + 0.7 m, described later, the ground subsidence was determined to be 1.6 m.

Figure 4(c) taken at St. 3 near the crematory further seaward of St. 2 shows the condition of the seawall. The seawater spouted out through a small hole of the seawall to the road on the east (right) side of the seawall of 0.8 m height. It is clearly understood from this photograph that the sea level was higher than the road level. Furthermore, Figure 4(d) taken at St. 4 shows the east side of the crematory, after walking straight along the road toward the sea. From this photograph, it seems that the crematory is floating on the sea with a mangrove forest.

Then, the situation in the vicinity of the crematory was observed in detail, which showed that almost all the land on the east side of the crematory subsided below the sea surface, and the trees were submerged below the sea surface, as shown in Figure 4(e). This crematory was located approximately 50 m landward of the shoreline in 2003, as shown in Figure 3(a), but the crematory sunk below the sea surface because of the ground subsidence and the sand discharge due to the westward longshore transport.

However, a low berm was formed near the retreated shoreline. Therefore, the berm height was measured relative to the sea level at 12:35 on 10 August 2022, which showed that the berm height was 0.6 m (Figure 4(f)). As the elevation of the sea surface at the measurement time was MSL + 0.10 m, the berm height relative to the MSL was 0.7 m. In addition, the vertical seawall protecting the seaward side of the crematory was severely damaged by wave action along with the inundation behind the seawall owing to wave overtopping (Figure 4(g)).

When the condition of a road extending along the east side of the crematory, as shown in Figure 3(e), was investigated after the observation near the crematory, the road was found to be protected against waves and erosion by both a 0.8-m-high vertical seawall and the zig-zag-shaped revetment in front of the seawall (Figure 4(h)). However, the seawall was severely destroyed by waves with no trace of its original shape remaining (Figure 4(i)).

Figure 5 shows an oblique image of the measurement carried out on July 1, 2022, using a UAV (drone: DJI Phantom 4 Pro V2.0). This image clearly shows that ground subsidence occurred in an extensive area along the road, and the ground on both sides of the road subsided below the sea surface. Although a slender mangrove forest remained west of the crematory, all the surrounding areas were submerged below the sea surface, and the seawall constructed along the coastline was invisible.

Figure 5.

Oblique image of the study area taken on 1 July 2022.

3.2 Subarea 4 (Sikebu River mouth)

The Sikebu River in subarea 4 flows into the Java Sea at a location 8.6 km west of the Semut River in subarea 3. The coastline inclination near this river mouth forms an angle of 37° clockwise relative to that of the mean coastline between subareas 1 and 3, as shown in Figure 1. Therefore, under the condition that waves are incident from the northeast, the angle between the direction normal to the shoreline and the wave direction becomes smaller than those in subareas 1–3, implying the decrease in longshore sand transport compared with those in subareas 1–3.

The satellite image taken in February 2005 is shown in Figure 6(a). At that time, the Sikebu River significantly meandered and then flowed out to the Java Sea near the river mouth because the mouth was closed by a sand spit extending southeastward. Here, wave crestlines can be clearly observed offshore of the shoreline in this satellite image. Therefore, the wave incident angle was determined to be N27°E from the direction normal to the crestline offshore of the shoreline. As the direction normal to the mean shoreline at that time was N50°E, as shown in Figure 6(a), waves were incident at an angle of 23° counterclockwise relative to the direction normal to the shoreline in this area, causing southeastward longshore sand transport, which is different from the predominance of westward longshore sand transport in subareas 1–3. The reversal in the direction of longshore sand transport from westward in subareas 1–3 to southwestward in subarea 4 is due to the change in the shoreline angle; the coastline near subarea 4 forms an angle of 37° clockwise relative to that of the mean coastline between subareas 1–3 because subarea 4 is located on the eastern side of the Comal River delta significantly protruding northward, and waves from the eastern side are incident counterclockwise relative to the direction normal to the shoreline.

Figure 6.

Satellite images of subarea 4 (Sikebu River mouth).

Figure 6(b) shows a satellite image taken in September 2009. The river, which meandered significantly southeastward, became straight because of the extension of two jetties at the river mouth. The south jetty was asymmetrically arranged with the south jetty being longer than the north jetty by approximately 70 m. Therefore, the shoreline may locally advance on the southeast side of the jetty owing to the wave diffraction effect of the jetty, but the shoreline formed was straight.

Then, until April 2017, the shoreline retreated in the area south of the jetty, whereas on the north side of the river mouth, sand was deposited, resulting in the shoreline reaching the tip of the north jetty (Figure 6(c), although the south jetty was shortened relative to that in 2009 (Figure 6(b)). The fact that the shoreline advanced on the north side of the jetty and retreated on the south side clearly indicates that southeastward longshore sand transport prevailed at this river mouth until 2017.

However, the shoreline on the north side of the river mouth was retreated by 80 m by June 2019, although the shoreline on the south side did not change (Figure 7(d)). Here, wave crestlines can be observed offshore of the left bank of the river mouth. Thus, the wave incident angle was determined from the satellite image; the angle of N59°E was determined from the direction normal to the crestline offshore of the shoreline, as shown in Figure 6(d). Thus, the wave angle was found to rotate clockwise by 32° relative to the initial wave angle (N27°E) in 2005 because of severe ground subsidence.

Figure 7.

Satellite images of subarea 5 (Comal River mouth).

Near the river mouth, many small roads separating fishponds disappeared owing to the ground subsidence. The same condition can be seen from the satellite image taken in July 2020, as shown in Figure 6(e). The shoreline on the north side of the river mouth markedly retreated because of significant ground subsidence west of the Sikebu River. When comparing the shoreline in April 2017 (Figure 6(c)) with that in July 2020 (Figure 6(e)), the shoreline recession on the north side of the river mouth reached 110 m until July 2021 because of marked ground subsidence. Thus, marked topographic changes occurred owing to the ground subsidence since 2019 at the Sikebu River mouth, causing the reversal in the direction of longshore sand transport, although southeastward longshore sand transport was predominant in 2005. It is concluded that when the non-uniform ground subsidence occurs along the coastline, the longshore inclination of the shoreline changes, resulting in the occurrence of longshore sand transport, even if waves are incident from the direction normal to the previous shoreline.

3.3 Subarea 5 (Comal River)

The shoreline changes in subarea 5 around the Comal River mouth were similarly investigated using the satellite images taken five times between September 2009 and May 2022 (Figure 7). The images show that this river flowed northward in September 2009 with an opening at the north end. It is found that the sediment supplied from the river was deposited around the river mouth, forming a shallow offshore terrace (Figure 7(a)). On the other hand, the amount of sand deposited on the west side of the river was smaller than that on the east of the river mouth, and an extensive lagoon enclosed by a slender barrier island was formed. The greater deposition of sand on the east side indicates that the sediment supplied from the Comal River was mainly transported to the east coast.

Then, in 2015, the right tributary was formed at a location 1.1 km south of the river mouth of the main stream (Figure 7(b)). At this time, an umbrella-shaped sandbar was remained offshore of the mouth of the main stream, and two slender sandbars have extended along the marginal line of the offshore terrace: one is westward and the other is southwestward. The effect of the main stream was assumed to be weakened because of the decrease in discharge, and the offshore terrace started to deform mainly owing to wave action.

In June 2017, although a shoal still remained offshore of the main stream, the size of the sandbar extending in the east–west direction offshore of the river mouth decreased, and the sandbar moved southward by approximately 120 m (Figure 7(c)). As the sediment was also supplied from the right tributary, another shoal started to be formed offshore of the mouth of this tributary. The formation of this shoal can be recognized from the formation of white-capped breaker lines along the external line of the shoal.

In July 2020, the slender sandbar located offshore of the mouth of the main stream moved by 400 m southward compared with its previous location in October 2015 because the amount of discharge of the right tributary increased (Figure 7(d)). At the same time, the increase in the size of the terrace offshore of the mouth of the right tributary can be seen from the distribution of the breaker zone. Then, until May 2022, the sandbar extending in the east–west direction offshore of the main stream further moved southward by approximately 180 m while maintaining its slender shape, and the river mouth of the main stream was completely buried with sand. Thus, since 2015, when the effect of the mainstream was weakened, a slender sandbar extending in the east–west direction offshore of the main stream has moved southward while maintaining its direction.

For the parallel movement of a slender sandbar while maintaining its shape, it is necessary that the predominant waves are incident from the direction normal to the shoreline of the slender sandbar. Therefore, the direction normal to the shoreline was determined from the satellite image (Figure 7(e)), which was found to be northward approximately parallel to the flow axis of the Comal River. Thus, the Comal River flowed northward until 2009 with the sediment being supplied north of the river mouth. However, because of the excavation of the right tributary by May 2022, all the discharge of the river flowed through this tributary. As a result, the mouth of the main stream was completely closed by a sandbar, and an offshore slender sandbar connected to the right bank of the river mouth was formed. The slender sandbar offshore of the river mouth moved southward owing to the action of waves incident from the north, while maintaining its shape.

3.4 Subarea 6 (Kauman River mouth)

The Kauman River mouth is located 5.2 km southwest of the main stream of the Comal River. In this area, the westward extension of the river mouth bar was investigated from the satellite images taken three times between June 2014 and July 2020 (Figure 8). In June 2014, the sandbar of the right bank extended by 69 m from point A at the west end of the mangrove forest (Figure 8(a)). Until August 2017, the sandbar elongated westward by 250 m west of A. During this period, the smooth sand supply from the right to the left bank was assumed to be interrupted owing to the extension of the sandbar on the right bank.

Figure 8.

Satellite images of subarea 6 (Kauman River mouth).

When a sand spit is formed at the tip of the barrier island, the sand spit itself develops owing to the continuous supply of sand from the upcoast, and almost all sand is used for the elongation of the sand spit, resulting in the disruption of continuous sand movement along the shoreline because of the discontinuity in shoreline at the tip. Noshi et al. [11] showed an example of rapid development of a recurved spit at the south end of Phan Rang Bay located in the southeast part of Vietnam, and the beach was eroded on the opposite side of the tip of the sand spit during the elongation of the sand spit. The same examples were observed along the curved shoreline north of Pengambengan fishing port located in the western part of the Bali Strait [12] and along the barrier island developed in Nakatsu tidal flat, Japan [13]. Accordingly, the same phenomenon is assumed to occur at the southeast end of the barrier island elongating along the Kauman River mouth.

Furthermore, in Figure 8(b), another sandbar started to elongate westward while covering the previous sandbar. This sand spit further elongated up to a location 520 m west of point A by July 2020 (Figure 8(c)). Here, in Figure 8(c), many wave crest lines can be seen offshore of the breaker zone; therefore, the wave angle can be determined from the shape of these crestlines; the wave direction indicated the north. On the other hand, when a line normal to the shoreline was drawn at the center of the long sand spit, the direction of this normal line became N42°W; waves were obliquely incident clockwise by 42° relative to the direction normal to the shoreline, and this condition is close to the angle of 45° at which the shoreline instability will occur owing to high-angle wave instability [14]. For this reason, intensive westward longshore current (sediment transport) was induced in subarea 6, resulting in the elongation of a sand spit.

3.5 Subarea No. 9 (Wirasa River)

The Wirasa River flows into the Java Sea at a location 3.8 km northwest of Pemalang. Figure 9(a) shows the satellite image of this river taken in October 2004. The parallel jetties were already constructed at the river mouth by October 2004, and the east jetty was longer than the west jetty. As the east jetty blocked longshore sand transport, the shoreline on the east side of the river advanced offshore relative to that on the west side, indicating the predominance of westward longshore sand transport. When drawing a line normal to the shoreline at a location 130 m east of the east jetty, the direction became N11°E. Similarly, the direction normal to the shoreline at a location 370 m east of the east jetty that was not affected by the blockage of longshore sand transport can be determined to be north. Finally, it is concluded that waves are incident at an angle of 11° clockwise at this river mouth. As the wave diffraction zone is formed west of the jetties against waves incident from the N11°E direction, a bay-shaped shoreline was formed immediately west of the west jetty.

Figure 9.

Satellite images of subarea 9 (Wirasa River mouth).

Similarly, Figure 9(b) shows the satellite image taken in September 2013. Because the east jetty was extended up to a location 220 m offshore of the seawall, the wave diffraction effect was intensified, causing the shoreline to advance by approximately 40 m west of the jetty. Here, wave crest lines can be seen at a location 130 m offshore of the east jetty; therefore, the wave incidence angle can be measured to be N16°E. This angle is close to that of N11°E obtained in Figure 9(a).

4. Longshore sand transport and shoreline instability

The incident wave angle in subareas 1–9 set along the shoreline of the Comal River delta was evaluated by two methods. First, the wave direction was determined from the direction normal to the wave crestline offshore of the breaker zone shown in satellite images (Type I). Second, the wave direction was determined from the difference in angle between the directions normal to the shoreline before and after the construction of a groin (Type II). Table 1 shows the wave angle determined in each subarea by Type I or II method.

Number of subarea	Wave direction	Wave angle	Wave angle	Date of taking satellite image
1	N36°E	22°	I	27 August 2013
2	N34°E	20°	II	19 April 2014 and 29 September 2021
3	N40°E	23°	II	4 May 2020
4	N27°E	−23°	I	1 February 2005
5	N	0°	II	May 2022
6	N	42°	I	5 July 2020
7	N	35°	I	5 July 2020
8	N12°E	22°	II	16 June 2017
9	N16°E	16°	I	16 October 2004

Table 1.

Wave angles at subareas 1–9.

Table 1 indicates that the mean wave angle in subareas 1–3 east of the Comal River delta was as large as 22° clockwise on average; therefore, marked westward longshore sand transport could occur. In contrast, in subarea 4, immediately east of the Comal River mouth, the wave angle was −23° in 2005 before the occurrence of the marked ground subsidence, which was the reversed direction relative to those in subareas 1–3. However, after the ground subsidence, the wave angle rotated by 30° in this area, causing the reversal in the direction of longshore sand transport. In addition, in subarea 5 at the tip of the river delta, the wave direction was from the north.

In subarea 6 west of the river delta, the wave angle was as large as 42°, which is sufficiently large to cause the development of shoreline undulation due to the mechanism of high-angle wave instability [14]. In domains, such as subarea 6, when a groin is extended to block longshore sand transport, a stable shoreline cannot be formed upcoast of the groin, and longshore sand transport may flow out downcoast turning around the tip of the groin [15, 16].

Uda and Serizawa [16] carried out a numerical simulation of the extension of a sand spit around groins constructed on a coast satisfying the condition of high-angle wave instability using the BG model [17], taking the coast of Narvacan in Lingayen Gulf in the Philippines. A model beach with a uniform profile ranging in the depth zone between the berm height (h_R) and the depth of the closure (h_c) was set up on the upstream end on a solid bed, as shown in Figure 10, with sand being supplied from the upcoast boundary between h_R and h_c, and subsequent beach changes due to wave action were predicted. A solid flat bed with the depth of h_c was assumed in the offshore zone. A seawall was set along the coastline concurrently with the installation of eight groins of 490 m length at 90 m intervals. Incident wave height was assumed to be 2 m, which is equal to the berm height, and wave period was assumed to be 8 s. Waves were assumed to be incident from the clockwise direction of θ = 55°. The wave field was calculated using the energy balance equation. The calculation was carried out up to 3 × 10⁴ steps, where 10⁴ steps correspond to 0.9 years. The other calculation conditions are shown in Table 2.

Figure 10.

Prediction of development of a sand spit around the tip of groins when waves are obliquely incident to the direction normal to the shoreline at an angle of 55° [16].

Wave conditions	Incident waves: HI = 2 m, T = 8 s, wave direction θI = 55° relative to x-axis
Berm height	hR = 2 m
Depth of closure	hc = 6 m
Equilibrium slope	tanβc = 1/10
Coefficients of sand transport	Coefficient of longshore sand transport Ks = 0.01; Coefficient of Ozasa and Brampton term K2 = 1.62Ks; Coefficient of cross-shore sand transport Kn = Ks
Mesh size	Δx = Δy = 10 m
Time intervals	Δt = 0.8 h
Duration of calculation	3.2 × 10^4 h (4 × 10^4 steps)
Boundary conditions	Shoreward and landward ends: qx = 0; Right and left boundaries: dqy /dy = 0; Lower right corner of beach edge: dqx /dx = 0 or dqy /dy = 0

Table 2.

Calculation conditions.

Figure 10 shows the results of the calculation. Because waves are obliquely incident from the clockwise direction at an angle of 55°, leftward longshore sand transport developed, and sand supplied from the northwest (right) boundary was deposited while forming a sand body, which moved southeastward over the tips of the groins. There is a gap between the tip of the groin and the downcoast shoreline at the groin located at the southeast end, so a slender sand spit elongated southeastward (leftward) while leaving a lagoon behind it. Then, a barrier island was formed after the elongation of the slender sand spit. These results corresponded well with the measured changes [16]. Thus, it is concluded that a statically stable shoreline cannot be formed using groins under high-angle wave instability, because longshore sand transport may flow out downcoast turning around the tip of the groin. It is required to maintain continuous sand movement on such a coast satisfying high-angle wave instability.

In areas further west, the wave angle in subarea 7 located 7.7 km southwest of the Comal River mouth is 35°, and those in subareas 8 and 9 are 22 and 16°, respectively. Although the oblique wave angle decreases with distance from the Comal River mouth, waves are obliquely incident clockwise at a large angle at any points, causing westward longshore sand transport. Therefore, when a jetty is extended on such a coast, careful attention on downcoast erosion is required.

5. Discussion

Along the shoreline of 33 km length between Pekalongan and Pemalang in Central Java, nine subareas were adopted from both sides of the Comal River delta, and the characteristics of longshore sand transport in each subarea were investigated using satellite images. As a result, it was found that the wave angle relative to the direction normal to the shoreline is large at 22° on average in subareas 1–3 east of the Comal River delta, causing significant beach changes due to westward longshore sand transport. In subarea 4 (Sikebu River mouth), the reversal of longshore sand transport was observed because of the marked ground subsidence.

Finally, the mechanism is schematically shown in Figure 11. When the wave angle is sufficiently small, the shoreline between the groins is stabilized for the shoreline to be normal to the wave direction (Figure 11(a)). However, when high-angle wave instability occurs, the volume of sand to be deposited upcoast of groin G₁ is minimal, and almost all the sand discharges downcoast and the shoreline retreats up to the seawall (Figure 11(b)), resulting in the formation of flying spits [14, 16].

Figure 11.

Schematic of shoreline changes around groins.

For a soft measure against erosion to be adopted on such a coast, the continuity of longshore sand transport must be maintained (Figure 11(c)). Thus, in subarea 6, the only way to maintain the present shoreline is to form a dynamically stable shoreline by maintaining the continuity of longshore sand transport, that is, sand bypassing.

Apart from beach changes caused by longshore sand transport, severe ground subsidence due to the excessive pumping of ground water occurred in subareas 1–4 west of Pekalongan with a maximum subsidence of the ground reaching approximately 1.6 m in height, and the submerged coastal land had been expanding. For the measures against the submergence of the ground and erosion by the spatial imbalance in longshore sand transport, the setting back of the seawall is under way at present together with the installation of pump stations to discharge seawater for the reduction in water level in subareas 1–4. Accordingly, not only short-period measures should be taken but also the pumping of the underground water, which is thought to be the primary cause of the ground subsidence, should be controlled in these areas. In addition to these, mangroves should be planted to form a forest in an extremely shallow sea offshore of the seawall to enhance wave attenuation. In the areas west of the Comal River mouth, the effect of ground subsidence was not very marked, but waves were obliquely incident at a large angle from the eastern side. Therefore, maintaining the continuity of longshore sand transport is required in the long-term methods such as sand bypassing.

6. Conclusions

It was found that, in the Pekalongan area, the height of ground subsidence reached approximately 1.6 m which is much greater than the expected sea level rise of 0.6 m in the coming 100 years reported in IPCC [18], and the sinking area has been rapidly expanding at present. Together with the ground subsidence, the predominant waves constantly developing under the easterly trade wind are incident to the coast, causing a significant shoreline recession downcoast of the structures. Along the coastline east of the Comal River mouth, the mean wave incidence angle was as large as 22° clockwise, whereas on the west side, the angle increased up to 42°, causing marked shoreline changes due to high-angle wave instability, resulting in significant beach changes. Therefore, in this area, maintaining the continuity of longshore sand transport is required in the long-term methods such as sand bypassing instead of structural defense using groins or detached breakwaters.

Acknowledgments

This study is part of a JICA study related to the Project for Planning Coastal Master Plan Against Beach Erosion on North Coasts of Java. We would like to thank JICA for the permission to present the results.