Tsunami and earthquake information for Indonesia and elsewhere - 2006/08/25 02:09Notes on earthquakes and tsunamis: (Source: USGS, and personal research):
• A Richter 6.8 quake south of Nias in May 2006 produced a small tsunami in Sibolga, and rattled Padang.
• By some accounts, the Telos and the Mentawai area is due for a major earthquake(s)- they have had them before and they will have them again-but the timing is completely unpredictable-it could be tomorrow, or it could be in 300 years. There were major earthquakes in the 1797, 1830s and possibly 1860s offshore of middle and southern Sumatra which produced major tsunamis and loss of life. The 1883 Krakatoa eruption also produced major tsunamis affecting southern Sumatra, the Mentawais, and Nias. (The Dutch colony at Lagundri Bay, Nias was wiped out by the 1883 tsunami, which came from Krakatoa). The average return time for quakes of large magnitude (say Richter 8.0) in this general region (not locally) is not definable, as the limited data only stretches back about 100-200 years, but at a rough guess of the order of a 20-50+ years or so (for roughly 8.0 earthquakes, similar to other megathrust coastlines), but this is still only an average, (Source: USGS, and personal research), and includes offshore quakes which may not be felt on land.
• Offshore earthquakes need to be around a Richter Scale of 7.0 or more before they can produce a destructive tsunami, although some bigger ones do not produce a tsunami at all, as it also depends on the rupture-style of the quake (megathrusts are the most dangerous, as opposed to pure horizontal slip), the exact location, the orientation of any underwater uplift, direction of movement etc etc.
• Earthquake destruction is worse in poorly constructed buildings, and where there is alluvium/sand rather than bedrock.
• Tsunamis vary in size and strength according to the shape of a coastline or embayment, and also in the size of successive waves. Sometimes the first wave is the biggest, sometimes up to the 4th or 5th wave is the biggest, often hours after the first. They also vary substantially in their length, from only a few km long for smaller tsunamis, to up to 1200km long (eg Aceh, 2004), depending on the area of underwater land that has been uplifted. The tsunami of Aceh in 2004 was unusually destructive partly because it was very high (up to 35m recorded), but also so unusually long-1200km!. Bigger offshore earthquakes tend to uplift larger areas of underwater land, thus producing bigger and longer tsunamis.
• If earthquake shaking is a long period, more drunken swaying/swooning, it means the quake is big but very far away, especially if it lasts for a long time. If the shaking is violent and jerky, obviously the quake epicentre is closer.
• A rough guide is as follows: if it shakes for a minute a 3m tsunami could be produced, if it shakes for 3 minutes a 5m tsunami could be produced, and if it shakes for 10 minutes a tsunami of 10m or more could be produced. (The shaking in northwestern Aceh in 2005 went for at least 10-15 minutes, and a tsunami height up to 35m was recorded in some locales).
• Boats at sea will usually feel a violent ‘knock’, similar to striking a reef, if a major earthquake occurs. This is from a shockwave traveling through the water, but there may be little or no sensation of ongoing shaking, as occurs on land. Water may also swirl and produce odd currents.
• If the water on the shore recedes rapidly, you may have up to 5-10 minutes before a tsunami arrives (as the trough of the wave has arrived first- eg Thailand in the 2004 earthquake), but some tsunamis exhibit no water recession at all (because the peak of the wave arrives first, eg Sri Lanka in the 2004 earthquake). If the earthquake is very close, you may have less than 5 minutes following water recession (eg Aceh Dec 2004, only about 30 seconds); if very far away, it may be longer than 5-10 minutes.
• Large earthquakes usually have no ‘cues’ which preceed their occurrence (eg December 2004 quake of northern Sumatra). They may in some cases be preceeded by other large quakes in the area, which produces a kind of cascading effect, in the order of days (eg Chilean quake of 1960). Build-ups of regional stress have occasionally also been predicted (Sumatran quake of March 2005). Large earthquakes always exhibit aftershocks, which steadily decline over days to months, but these aftershocks may also be large, or more rarely, larger than the first shock.
• If a region of plate collision with a known history of large earthquakes has not exhibited any moderate to large quakes over periods of centuries, it is likely the area will, in a period of decades to centuries, exhibit large earthquakes. ‘Return periods’ (average period between major earthquake episodes) are suggested below for some regions of the world, although these are only rough averages, from limited data over the last few hundred years (Mediterranean Europe and China longer). The figures are based on such things as recorded historical quakes, calculated and observed rates of plate movement, direction of movement, plate and plate collision type, tectonic orientations, formation of regional stresses etc etc.
EG for Richter 8.0 earthquakes, return periods are of the order: (Note: some recent large earthquakes magnitudes and dates are also given (approx.)), and figures are for broad areas (eg whole of Japan), not local areas (eg Tokyo), and include areas immediately offshore):
o Southern Chile 20-50 years, (9.5, 1960) o Middle Chile 20-50 years, o Northern Chile 20-50 years, Japan 20-50 years, (5 8.0 earthquakes or greater in the last 100 years) o Kamchatka 20-50 years, (9.0, 1957) o Southern Alaska 20-50 years, (9.2, 1964) o Offshore western Alaska 20-50 years o Southern Sumatra 20-50 years (8.0, 2000, 1797, 1830s), o Northern Sumatra 20-50 years, (9.1, 2004; 8.7, 2005) Southern Peru 20-50 years (Estimated 8.7?, 1786?), o Northern Peru 20-50 years, Bali to Nusa Tengarra (East Indonesia) 20-50 years, o Java 20-50 years (7.7, 2006), o Andaman-Nicobar Islands 50-100 years (9.1, 2004) o Northern New Guinea 50-100 years, o Western USA 50-100 years, (7.8?, 1906) o Mexico, 50-100 years (8.1?, 1992?), o Ecuador 50-100 years, (Estimated 8.8?, 1906) o Northern Pakistan, 100 years, o Southwestern China, 100 years, o Central America 100 years, o New Zealand, 100 years, Far western Canada 100 years (Estimated 9.0, 1700), Phillippines, 100 years, o Caribbean and northern South America 100 years, Northern India and Nepal/Tibet, 100 years, o Turkey, 200 years, o Iran, 200 years, o Taiwan and southeastern China 200 years, o Mediterranean Basin 300 years, (Estimated 8.5?, 365 AD) o Portugal and Southern Spain 300 years. (Estimated 8.7-9.0?, 1756)
• For magnitudes of Richter 7.5, the periods given above are about 80% shorter, and for a Richter 7.0 magnitude the period is about 90% shorter. For example, in Java and immediately offshore a magnitude 8.0 earthquake can be expected about once every 20-50 years, a 7.5 occurs about once every 5-10 years, and a 7.0 occurs about once every 5 years. (This is for the whole island and offshore, not local areas. For a local effect, the return period would be siginificantly greater (say ten times these figures)).
• For extremely large (>=Richter 9.0) earthquakes, return periods are of the order of 5 times the above, but these generally only occur in the most active plate collision zones. Examples include Chile, Peru, Ecuador, Alaska, Kamchatka, and Indonesia; and possibly Japan, western Canada, southern China, and offshore Portugal.
• There were only 5 earthquakes in the world in the last 100 years of 9.0 or greater (southern Chile 9.5, southern Alaska 9.2, Western Alaska?, Kamchatka 9.0, and Northern Sumatra 9.1), about 40 of 8.0 or greater, and about 1700 of 7.0 or greater.
• The 9.1 event in northern Sumatra was, very roughly, a 1 in 250-500+ year event for that area. (Note the 8.7 event can be regarded as an aftershock of the same episode, depending on who you ask). The estimated 8.7-9.0? earthquake in Portugal in 1756 appears to be around a 1 in 1,000-2,000 year event. There has not been an earthquake of equivalent size in this area in the last 1,000 years at least.
• Note also the Richter scale is logarithmic, a 7.5 is about 5 times smaller than an 8, and a 7.0 is about 32 times smaller than an 8.0.).
• Tsunami frequency: (note: only earthquakes that are offshore or nearshore and greater than about Richter 7.0 can produce a shoreline tsunami height greater than about 1m, and only about 50% of these actually do so. The proportion is higher for offshore earthquakes greater than 8.0, which also tend to produce larger tsunamis. Shoreline tsunami height is also affected by local embayments etc. Also, very large earthquakes with epicentres on land, but near to coastlines, can also produce tsunamis, because the area of land that can be uplifted extends over the water. A confounding factor is tsunamis can produce underwater landslides which can increase tsunami size, especially for smaller quakes around 7.0).
o A moderate tsunami of 1-2m height occurs about 5-10+ times per year, (may be a minor number of fatalities). These may be localised and not recorded, produced by less than 50% of offshore or nearshore 7.0+ magnitude earthquakes. o A destructive tsunami of 2-5m height occurs 1-2 times per year (may be hundreds of fatalities or more), produced by less than 50% of offshore or nearshore 7.2+ magnitude earthquakes. (eg Java 7.7 2006, about 3m; Nias 8.7, 2005, about 3m) o A very destructive tsunami of 5-10m height occurs about once every 5 years, (may be thousands of fatalities or more), produced by less than 50% of offshore or nearshore 7.6+ earthquakes, and o An extremely destructive tsunami of >10m height occurs about once every 10+ years (may be tens of thousands of fatalities or more), produced by less than 50% of offshore or nearshore earthquakes greater than ~8.2 magnitude (eg Aceh 2004, 9.1, up to 35m high, Peru 1700s, 8.7? up to 25m high).
Tsunamis greater than about 20m might occur about once every 50+ years, which might need earthquakes greater than about 8.7+ magnitude.
Note that some smaller tsunamis occur in relatively unpopulated and localised areas (maybe only over a few kms or so), and may not be recorded by scientific instruments at all. (However the seismic waves from the earthquake are always recorded). Such localised tsunamis may also not produce any fatalities, or both the tsunami and any fatalities may not be documented.
• The worst places for tsunamis in the world are in offshore megathrust zones, such as Chile, Japan, southern and western Indonesia, Andaman-Nicobar islands, northern parts of New Guinea, Kamchatka, southern Alaska, Peru, and Ecuador; and to a lesser extent Central America, the Carribean, the Mediterranean, offshore Portugal, far western Canada and USA, the Phillipines, Taiwan and southeastern China, and parts of the Southwest Pacific. Tsunamis generated in these regions can also extend to other regions.
• A few rare intra-continental zones are prone to large earthquakes such as Missouri USA.
• Some Historical tsunamis (many others):
365 AD, Mediterranean, Greece, Italy, Egypt, several 10,000s of fatalities. 1755 Portugal and Morocco, several 10,000s of fatalities. 1830s Southern Chile, several 1,000s of fatalities. 2004 Northern Sumatra, India, Sri Lanka, Thailand, several 100,000s of fatalities. 1960 Chile, 9.5 Japan, several 100s fatalities 1964 Alaska, California, 9.2 several 10s of fatalities 1700 Western Canada, ?fatalities 1908 Italy, 10,000 fatalities 1883 Krakatoa (volcanic), Nias, Mentawai, Southern Sumatra, several 10,000s of fatalities. ~~BC Santorini (volcanic), Greece, 1,000s of fatalities. 2006 Java, 7.7 several 100s of fatalaties Migrated message from the old Wannasurf forum. Author: ozsurfer (rogermcevilly at hotmail dot com)