The September 9, 2002
Wewak, Papua New Guinea
Earthquake and Tsunami

The information presented in this web page was collected on several surveys performed between September 9th and October 1st, 2002. Contributors to this report are: Professor Hugh Davies, UPNG, Christine Saiang, UPNG, James Bu, UPNG, Burak Uslu, USC, Professor Costas Synolakis, USC, Jose Borrero, USC, John Freckman, California State Lands Commision, Brandon Gomer, Northwestern University, and Professor Emile Okal of Northwestern University.

Introduction

At 18:44 UTC on September 8, 2002 (04:44 on Sept. 9, local time), a strong earthquake (mb = 6.5, MS = 7.8) occurred off the northern coast of the East Sepik Province of Papua New Guinea ("PNG"), and triggered a moderate tsunami which was observed along 250 km of the northern PNG coast. This event is of particular interest since it occurred only 120 km ESE from the earthquake of July 17, 1998 (mb = 5.9, MS = 7.0), which triggered a devastating tsunami with a death toll exceeding 2100.

Papua New Guinea

Earthquake Details

Hypocentral parameters for PNG02 are 3.30S; 142.95E, and a depth of 13 km, corresponding to a location approximately half-way between Aitape and Wewak and within a few km of the shoreline. Focal mechanisms available in the immediate aftermath of the earthquake consisted of Harvard's "QUICK CMT" whose final solution is strike =106; dip = 34; slip = 43. The CMT moment (2.97 x 1027 dyn-cm) suggests a fault length L = 72 km, a width W = 36 km, and a slip u = 2.13 m. Using comparable parameters, Matsuyama (pers. comm., 2002) has suggested that observed crustal deformations (in particular uplift on the offshore islands of Walis, Tarawai, Kairiru and Mushu) show better agreement with the shallow dipping fault plane mechanism than with its steeply dipping conjugate.

Initial Surveys

Immediately after the event, one of us visited the area in the company of senior government officers to assess damage, identify areas that might be at further risk due to unstable ground, discuss the situation with local authorities and people in the affected villages, and report back to the central government. A second visit followed a week later after several strong aftershocks had caused further concern (Davies, 2002; Davies et al., 2002). These early surveys were instrumental in focusing and guiding the tsunami survey team which arrived two and a half weeks after the event.

The preliminary surveys showed that it was the earthquake, rather than the tsunami, that caused most damage. The shaking was responsible for landslides, liquefaction, collapsed buildings and water tanks, a severed water supply main for the town of Wewak and, remarkably, the uplift of the outer islands by 30-40 cm. The earthquake damage was more severe in areas underlain by unconsolidated sediment. Several people were injured and five killed when houses collapsed or as they hurriedly escaped from shaking buildings. Houses built of traditional materials can be repaired without significant external assistance. However a more critical effect was the loss of many water tanks which toppled during the shaking.

The earthquake caused small extensional faults to develop in areas of unconsolidated sediment, notably at Kauk beach and at Ubidnim, near the mouth of the Hawain River. On the east and north coasts of Kairiru Island many landslides were triggered, and fissures with displacement of 10-30 cm opened on steep slopes as a result of incipient slope failure. Liquefaction of unconsolidated subsurface sediments was common at localities from Kauk eastwards and was most pronounced at Ubidnim, where large blowout holes developed and water and sand spouted 3-5 m into the air.

The Northern Coast of Papua New Guinea.

At Ubidnim and other villages, liquefied sediment moved laterally and upwards to completely fill open water wells. The earthquake caused the collapse of brick school buildings at Hawain but otherwise caused little damage to western-style buildings. In the villages, about 10% of the houses, built of bush materials, collapsed completely and another 20% were damaged to some degree; however the severity of damage varied from place to place. Residents reported that the earthquake was preceded by, and accompanied by, a roaring noise that approached Wewak from the west-northwest. Meanwhile, night fishermen reported a bright blue-green glow in the sky above Kreer, immediately south of Wewak.

Road approaches at several bridges subsided by 10-20 cm, and a landslide temporarily blocked the coast highway at Hawain. In Wewak, the water main was severed in two places. The first occurred where a lightweight frame carrying the pipe across a river collapsed. The movement of unconsolidated material around a buried section of the concrete pipe caused the second failure.

The outer islands: Tarawai, Walis, Kairiru and Mushu, were elevated by 30-40 cm. At St John's Seminary on the south coast of Kairiru Island, 35 cm of uplift was measured with tide marks on a small pier. Eyewitnesses on Tarawai reported that the uplift happened some time after the earthquake, in two stages an hour apart. The uplift caused the emergence, at mid-tide, of reefs and wave-cut platforms, and the setting of new strand lines on beaches. As a result of the uplift, swamp lands on Tarawai, Walis and Mushu have been partly drained, and water in wells has dropped to a lower level. This problem may become more severe as the swamplands are important as a source of sago, a staple food, at times when garden produce is in short supply.

Tsunami Survey

The highest runup measured (~5 m) was on the south side of Victoria Bay on Kairiru Island. High runup values were concentrated on the west-facing bays of Kairiru and Mushu Island and on the mainland to the south and west. Runup and inundation tapered off gradually to the east and west of these centralized maxima. Due to the possible amplification of runup heights in the western facing bays of the offshore islands (Kairiru and Mushu) the field data in Figure 4 have been divided into mainland data points and island data points.

Western Region
To the west of Aitape, there was very little tsunami activity. Residents in the area reported feeling the earthquake very strongly but generally did not notice any extreme water motions. At Sissano Lagoon, a site which was completely devastated in the 1998 tsunami, one fisherman reported strong surges flowing through the mouth of the lagoon. These surges persisted for some hours after the earthquake. To the east at Malol, the residents reported that, having fled the coast after the earthquake, upon returning they noted a uniform line of small coral pebbles deposited along the beach. This line was later washed away by high tide, however some were still present at the time of the survey 2 weeks after the event.

Photos from Sissano (top row) and Aitape.

At Aitape, the tsunami was noticed as a series of surges that flowed up the Aitape River and spilled over the banks and flooded some residents' gardens. Many distinct surges were reported at Aitape, with the first and largest coming 30 minutes after the earthquake. Residents estimate that there were as many as 12 surges continuing for up to 2 hours after the earthquake. The earlier surges were the largest and caused the inundation over the riverbanks. The Eastern Shore of Seleo Island offshore of Aitape was also visited. Residents reported unusually low water levels for up to 1.5 hours after the earthquake followed by a maximum positive surge of 1.2 m. Further east on the mainland, residents in the village of Paup reported that their canoes were moved by a surge some 30 minutes after the earthquake. They reported up to two surges reaching over 2 m above sealevel. This is in contrast to the next site eastward on the mainland, Suain, where the residents reported no positive surge, however they did note that the ambient tide level seemed 'lower than normal' for several days after the earthquake.

Central Mainland Region
The central region from Kauk to Cape Moem, including the offshore islands, was the area most affected by the tsunami. On the mainland, the village of Kauk experienced a positive surge of nearly 1.8 m while 15 km to the east at But, there was no wave activity at all. The villages of Karawop, Boiken, Krupier and Yuo were all affected by the tsunami. Of the three, Boiken, situated on the eastern side of a small headland was the hardest hit. Three distinct surges were reported. The first came in uniformly from the sea while the second came from the west, sweeping across the exposed portion of the headland and along the shoreline through the village. Two individuals were entrained by the flow and nearly drowned. Maximum runup was measured at over 2 m while inundation distances varied between 75 and 190 m.

Paup, But, and Kauk.

At Krupier, east of Boiken in the bay formed by the Boiken headland, the surge came in directly from the north, overtopped the beach berm flowed through the village and into the swamp behind it. Flow depths in the village were estimated between 40 to 60 cm, lower than the floors of the local stilted huts, i.e., the material inside the huts was not wetted by the wave. Inundation of 80 m was measured to the edge of the swamp, however this is a minimum value since the surge continued inland through the dense swamp.

Maps and photos from Boiken.

Slightly smaller runup was measured to the west at Karawop and to the east at Yuo Village at the mouth of the Hawain River. It should be noted that Yuo Village is situated 10 km east of Boiken on a headland very similar in size and shape to the Boiken headland, however the tsunami was much smaller and less energetic than at Boiken.

Keresau, Walis and Tarawai
Offshore of this area are the islands of Keresau, Walis and Tarawai (Figure 2). Keresau, which lies closer to the mainland, directly offshore of the Boiken headland, experienced some flooding and positive runup. Three locations were measured on the southern shore of Keresau facing the mainland. Here residents observed the wave flooding the village grounds. A clear debris line was also observed. Runup was measured between 0.9 and 1.1 m with inundation distances of 10 to 30 m. Along the north shore of Keresau, there were no eyewitnesses and there was no reliable indication of a positive wave.

Photos from Walis and Tarawai.

Further offshore on Walis and Tarawai islands, residents only reported a withdrawal of the sea with no positive wave: shortly after the earthquake the sea withdrew and stayed out for many hours. The reefs were exposed some 500 m offshore. The islands of Tarawai and Walis have been clearly uplifted. Rough estimates of uplift on Tarawai were taken by comparing the current high tide (measurements happened to be taken at high tide) to the levels of other high tide marks. Our measurements yielded an uplift of 50 cm on the eastern end of Tarawai Island. Evidence of uplift was also observed on Walis Island. On this particular day at high tide, the sea level was observed to be 30 to 50 cm less than the level normally expected for the high tide, as indicated to the ITST by local residents on a log bridge and on the root outcrop of a tree growing at the shoreline.

Mushu and Kairiru Islands
The two large islands of Kairiru and Mushu which lie to the northwest of Wewak were greatly affected by the earthquake and tsunami. Kairiru is a rugged steep island with a maximum elevation of 760 m while Mushu is much flatter with elevations only reaching 120 m. Both islands have west-facing bays that seemed to have trapped the incoming tsunami wave and amplified the runup. The largest runup was measured on the south side of Victoria Bay on Kairiru Island. A clear water mark and debris line was measured at 5.5 m. Also in Victoria Bay the village of Shauer on the north side of the bay was inundated by the wave. The wave penetrated over 100 m and reached an elevation of 2.1 m. At the head of Victoria Bay, there was evidence of overtopping of the berm and of inundation up the small river.

Photos from Kairiru.

;

;
; ; ;

On Mushu, the village of Warak was inundated to a distance of 100 m and a maximum runup of 1.5 m. Two structures were damaged by the tsunami in Warak and there was extensive evidence of tsunami sand deposition. A uniform layer of white beach sand approximately 1 ­ 2 cm thick was overlain atop a layer of thick black mud. There was also evidence of the flow depth at Warak given by watermarks on the Sago Palm trees at the back of the village some 70 m inland.

Tsunami deposits at Warak on Mushu Island

East to Wewak
Along the mainland east to Wewak, a village on the Wom Peninsula was inundated. The inundation was 40 m with a runup of 0.66 m. Just to the west of Wom, at Cape Pus and the entrance to the Pus Lagoon, one resident, who had gone there to fish several hours after the earthquake, reported seeing surges flowing in and out of the lagoon. He also noted that water was flowing out of cracks in the ground near shore.

In Wewak, three locations were surveyed. Along the main beach at a petroleum depot, a night watchman reported that a surge flowed over the seawall and on to the coastal road. There were also reports of unusual surges flowing up a drainage canal, for up to few hours after the earthquake, as also observed at Aitape and Cape Pus. East of Wewak at Moem, there were also reports of a tsunami after the earthquake. Several locations were surveyed and runup was measured at approximately 1.2 m. There was evidence of sand deposition from the tsunami overtopping a sea wall and flooding across a coastal road. No data was taken between Moem and the Murik Lakes area due to accessibility and time constraints.

Murik Lakes
In the far eastern region of the survey area, the Murik Lakes area features a series of lagoons and barrier islands with very low topographic relief. Local residents say that their land is regularly flooded by strong ocean swells. Three locations were surveyed which all gave similar accounts of the sea rising above the berm (~ 1 m) while the lagoon level also rose, flooding the villages from inland, to a depth of 30 cm to 1 m.

Pictures from Murik Lakes. the narrow sand spit was over topped with flooding coming from both the sea and the lagoon.

Data Summary

The following table summarizes the maximum runup data collected in the field. The points are differentiated between mainland and offshore island points. The west facing bays on Mushu and Kairiru Islands may have amplified the runup.

References

  • Davies, H.L., Field Survey Report of the September 9, 2002 Wewak Earthquake, International Tsunami Information Center Newsletter, October 2002.
  • Davies, H.L., and the Wewak Survey Team, Preliminary results of the field survey, 2002 PNG earthquake at Wewak, Eos, Trans. Amer. Geophys. Un., 83, (47), F1005, 2002 [abstract].
  • Matsuyama, M., Preliminary Numerical Simulations of 2002 Wewak earthquake tsunami. Sept. 19, 2002.

Counter