An International Tsunami Survey Team (ITST) studying the effects of the December 26 tsunami on Indonesia's island of Sumatra documented wave heights of 20 to 30 m (65 to 100 ft) at the island's northwest end and found evidence suggesting that wave heights may have ranged from 15 to 30 m (50 to 100 ft) along a 100-km (60-mi) stretch of the northwest coast. These wave heights are higher than those predicted by computer models made soon after the earthquake that triggered the tsunami. "Groundtruthing" the models, which are used to forecast tsunamis for early-warning systems and long-term planning efforts, was one of the main goals of the scientific survey.
The survey was conducted from January 20 to 29 in the province of Aceh, which lies only 100 km (60 mi) from the epicenter of the earthquake and sustained what many consider the worst tsunami damage of all affected areas. About a third of the 320,000 residents of Aceh's capital, Banda Aceh, are dead or missing, accounting for much of Indonesia's toll of more than125,000 dead and 90,000 missing.
Led by Yoshinobu Tsuji of the University of Tokyo's Earthquake Research Institute, the survey team consisted of nine scientists from Japan, six from Indonesia, two from France, and two from the United States. The U.S. scientists were Andy Moore of Kent State University and Guy Gelfenbaum of the U.S. Geological Survey (USGS).
The team collected information about wave heights at the beach and inland, inundation distance (how far inland the water reached), runup elevation (the water's height relative to mean sea level at its farthest reach inland), flow directions, erosion, sediment deposition, and coastal subsidence.
The team gathered some of its information from eyewitness accounts. Though not always reliable (commonly, eyewitnesses are running for their lives as they observe the tsunami), the eyewitness accounts collected in Sumatra provided several consistent pieces of information:
Being just landward of the subduction zone where the tsunami-generating earthquake occurred, northwestern Sumatra was struck by a "near field" tsunami. In contrast, areas across the ocean from the earthquake epicenter, such as Sri Lanka, were struck by "far field" tsunamis. The rapid arrival of the tsunami in near-field locationsjust 15 to 20 minutes in Banda Acehmeans that a tsunami early-warning system, now in the planning stages for the Indian Ocean, should be accompanied by tsunami education and long-term emergency and land-use planning efforts for the most effective mitigation of tsunami hazards.
Because the tsunami washed out many roads and bridges, the scientists had to hike long distances to reach some field areas, and on several occasions used makeshift rafts constructed from barrels and boards to cross rivers. Despite such complications, they were able to collect much data, which will be used to improve both the scientific understanding of tsunamis and the computer models used to predict tsunami effects.
To measure wave heights, the scientists looked for water stains on buildings and broken branches and debris in trees (where buildings and trees were left standing), then used laser rangefinders to calculate the heights. The west-facing coastlines were struck by the highest waves, some more than 30 m high. Waves that wrapped around the island to hit the north-facing coastline of Banda Aceh were lower, about 10 m high, but the area's low-lying land allowed those waves to penetrate far inland. Inundation distances in the province were so large that they were most easily measured from satellite images, where sediment deposited by the waves and vegetation killed by the saltwater are clearly visible. One such image shows that the waves that struck the villages of Lampuuk and Lho Nga on the west coast met the waves that struck Banda Aceh from the north.
Items broken and bent by the tsunami waves, as well as sedimentary structures in the tsunami deposits, were used to determine flow directions. The scientists found that the large tsunami waves flowed around natural barriers, flooding low-lying areas behind them.
The researchers surveyed beach profiles to document erosion (common near the coast) and deposition (common inland) by the tsunami. Sand eroded from beaches probably provided much of the sand that was deposited inland. The survey team dug trenches in the tsunami deposits to measure their thickness and to examine other characteristics that can shed light on how high the waves were and how fast the water was flowing. Data from the sediment deposits will not only tell scientists about the recent tsunami but also help them recognize and interpret the deposits of ancient tsunamis, which, in turn, will help them better understand an area's tsunami history and its likely tsunami risk (see "USGS Scientists Study Sediment Deposited by 2004 Indian Ocean Tsunami" in Sound Waves, February 2005).
Models predict that the type of earthquake that caused the tsunamia megathrustwill raise the sea floor above the fault rupture and cause subsidence near the coast. So, the team was not surprised to find evidence that coastal land had subsided in Sumatra. Trees with roots and lower trunks submerged in seawater indicate that coastal land subsided 1 to 2 m (3 to 6 ft) in some areas. Japanese team member Yuichiro Tanioka and his Indonesian graduate student Yudhicara resurveyed parts of Banda Aceh for which older elevation maps were available and discovered that the land there had subsided by 28 to 57 cm (about 1 to 2 ft).
The scientists' efforts were focused mainly around the very northwest end of Sumatra, but they also collected data about 100 km (60 mi) to the south, at Kreung Sabe. Wave heights of 15 m (50 ft) at that site suggest that the tsunami waves may have been unusually high, 15 to 30 m, along the entire 100-km stretch of coast from Kreung Sabe to the northwest tip of the island. USGS scientists hope to return to Sumatra in April to test this hypothesis by measuring wave heights at intermediate points along the coastline and to collect additional data, such as nearshore bathymetry and sediment-deposit profiles.
in this issue: