Knowing versus Doing

Go to: "Disaster Preparedness and Response: Why is the Phone Off the Hook?" by Ben Wisner. Invited paper for the European Telecommunications Resilience & Recovery Association Inaugural Conference (ETR2A), Newcastle-upon-Tyne, UK, 11-13 June 2003

 Go to: "Same errors recur, despite quake lessons learnt" by David Alexander. AlertNet http://www.alertnet.org/thefacts/reliefresources/alexanderview.htm

 Go to: AlertNet - UNU social vulnerability project dissemination via videos: http://www.alertnet.org/thefacts/reliefresources/384095

Below:

·        Cinna Lomnitz Why did they have to die?: No matter how terrible the earthquake, buildings could survive - if built properly in the first place. 

·         Ben Wisner 

·        Haresh Shah recalls his 1999 email discussion

·        David Alexander Notes on landslide hazard and risk identification and remediation 

Why did they have to die?: No matter how terrible the earthquake, buildings could survive - if built properly in the first place. 

Cinna Lomnitz considers the lessons of catastrophe and the challenge to engineers.
[24 Jun 1999: The Guardian Page 1]

In the 1985 Mexico earthquake about 400 modern reinforced concrete-frame buildings between seven and 18 storeys high were destroyed, with up to 472 people killed in one building. It was the worst in the long history of disasters to have struck Mexico. Comparable seismic disasters were quite unknown in pre-Hispanic America. The Aztecs lacked the benefits of modern engineering but their lives were safer than ours. When Professor James Brune showed some years ago that there is a natural upper limit to seismic ground accelerations, the Aztecs would just have nodded.

When they felt challenged by nature they tended to do something about it.

The housing standards of Mexico City, as the Spanish sighted it in 1519, were far superior to anything then to be found in Europe. The city was larger and better planned than Madrid. Three causeways linked it to the mainland: the causeways had removable bridges which made a surprise attack all but impossible. The population was self-reliant as all their needs were supplied by the lake and by the mud which they hauled up from the lake to construct their rectangular chinampas or hydroponic fields.

Most life forms tend to adapt to their environment - and earthquakes are part of our environment. Are we the exception to the rules of creation? It is hard to find another species besides Homo sapiens which has not learned to make their nests or burrows safe against earthquakes. It is good to remember that no ground motion as large as 2g has ever been recorded in an earthquake. But we may easily simulate an acceleration of this magnitude. Suppose we are riding at 70 km/h and we apply the brakes to bring our car to a standstill within 20 metres: this test has subjected us and our automobile to a deceleration of about 2 g. Any car or human should be able to survive this force without damage. Why can't buildings?

As a 14-year old boy I lived through the great Chile earthquake of 1939, magnitude 8.3, which killed 28,000 people. Some years later, another catastrophic event surprised me as I was walking along the fence of Concepcion Airport, some 200km from the epicentre of the 1960 earthquake of magnitude 9.1- believed to be the largest earthquake in human memory.

I felt the 1952 Kern County, California earthquake, magnitude 7.4, while I was a student in Pasadena under Gutenberg and Richter. I was in Caracas during the 1967 earthquake, magnitude 7.1 which destroyed a number of 12-storey reinforced concrete-frame apartment houses sited on soft ground.

I lived through the 1985 Mexico earthquake, magnitude 8.1. It is enough to make me wish not to have to live through another earthquake. On the other hand, there is something to be said for having felt damaging earthquakes at first hand.

The human sensory apparatus, imperfect as it may seem to us, remains superior to commercial seismic instrumentation in some relevant aspects. One type of motion accelerates the structure in the same direction into which it tilts. This prograde motion is familiar to anyone who has tried to stand on the deck of a ship. A wave on the surface of a liquid is prograde, and causes any floating object to move in this fashion. The opposite type of motion is known as retrograde it impels the object in one direction and tilts it in the opposite direction.

Charles Darwin described feeling prograde ground motion in the great 1835 Chile earthquake. He wrote: 'The motion made me almost giddy: it was something like the movement of a vessel in a little cross-ripple, or still more like that felt by a person skating over thin ice which bends under the weight of his body.' I too have felt earthquakes on soft ground. I know exactly what Darwin meant and so did the structures in the recent Mexico, Loma Prieta and Kobe earthquakes before they collapsed. We shall see that those three recent disasters have more than a little in common, and there is some evidence that prograde ground motion could have something to do with it.

Ocean liners are damped, to control rolling and to prevent capsizing. Cars are damped to control the effect of rough or uneven pavements. Most cars can stand several g's of horizontal acceleration - more than the most severe possible earthquake can produce. I am told that no building code in use today could have prevented the disastrous collapses of modern structures in Kobe or in Oakland.

On the other hand, every single failure was explained in terms of inadequate design or workmanship. If our building fails it will be our fault. Those structures in Kobe, Oakland or Mexico City were not made for a soft-ground environment. Charles Darwin could have told you that. How come our instruments cannot record what Darwin felt? The answer should amaze no one: we build our instruments to suit our ideas of what we expect to record. We do not expect the rotational components of ground motion to be particularly important so we don't record them. Nature thought otherwise or she would not have endowed us with special sensors for rotational accelerations. These sensors are placed strategically on either side of the head, in the inner ear. They enable us to feel all three components of rotations and to tell prograde from retrograde ground motion, as Darwin clearly could. He was sensitive to the difference because he was prone to sea-sickness, which means that his inner ear was attuned to prograde motion as found in ocean waves. Sea-sickness is a defensive response of the nervous system to prograde ground motion.

Consider now norms called 'building codes', and specifically the earthquake norms. Civil engineering structures such as buildings, bridges or dams are unique industrial products: they are rarely if ever tested before use.

Detergents, cars, computers, glassware, medicines, even military weapons are tested under realistic conditions. It seems to me that we don't try hard enough. The potential loss of life is far greater in the case of bridges, dams or buildings than it is in the case of a bomber or a submarine. Soldiers go to war fully expecting to be shot at. Civilian families who purchase an apartment in a reinforced concrete-frame structure because it is advertised as earthquake-resistant are in a different position.

I can still hear the cries of small children buried alive under the collapsed concrete structures of Mexico City. There were hundreds of them. In most cases there was no way to reach them safely and they perished alone in the dark. Sometimes the dying took hours or days. We have a right to be revolted by such a prospect. Is it the fault of the building codes, or the absence of standardisation, or the lack of inspection or quality control, or what?

According to statistics, 16.7% of some 2,000 high-rise structures collapsed in downtown Mexico City during the 1985 earthquake. Earthquake damage is not God-ordained. It is cultural in origin. The theoretical peak acceleration in earthquakes - 2g - has never been actually recorded. There are mass-produced structures, such as automobiles, that can survive an acceleration of 2g without damage. Therefore it would seem that total earthquake safety is a reasonable aim. In the 17th and 18th centuries some Mexican Colonial engineers started to build churches and other large buildings as if they were ships. Many of these structures successfully survived large earthquakes, including the 1985 earthquake. But the art of building structures as if they were ships has apparently been lost.

Cinna Lomnitz is a professor of seismology and earthquake engineering at the National University of Mexico. This article is based on a lecture in London in May at the Institution of Civil Engineers.

Ben Wisner, original RADIX 'think-piece' 

"As of this morning 39 % of hospitals were lost and significant damage to laboratory materials has been discovered. Temporary clinics are being set up to in several areas to accommodate the fact that over 1300 beds have been rendered unusable by damages caused by the earthquake. In the case of the Rosales and San Rafael hospitals (see detailed report on the web) damages are structural and will take years to be fixed." [PAHO, 19 January 2001, see http://www.paho.org]

Since the 1985 earthquake in Mexico City, PAHO has worked hard to develop with partners in the region methods of structural and non-structural mitigation for hospitals. These have been published and are available free of charge. And PAHO experts have been available to assist in implementing this knowledge, knowledge that is well established. Why hasn't it been used? Why in the year 2001 does this earthquake knock out 39% of El Salvador's hospitals.

"As usual in Central America, solidarity displayed by neighboring countries resulted in offers of assistance and the immediate sending of health professionals, equipment and mobile facilities. Although these mobile facilities are no substitute for normally operating health services, they are sufficient to respond to life saving needs. No medical teams, in addition to those from neighboring Latin American countries and the U.S. Military (SOUTHCOM) are likely to be required.

As a policy, PAHO/WHO discourages sending mobile field hospitals from other than the closest of geographical neighbors sharing the same culture and health approach, because they are costly, difficult to transport and arrive too late to make a difference in terms of saving lives. The high cost of this type of aid (which also quickly depletes the donor's budget) would be better invested in medium-term needs that often go unmet once public attention wanes." [PAHO, 14 Jan. 2001, my emphasis]

How well is this policy generally accepted- Implemented? Does the U.S. Military share "the same culture and health approach" in PAHO's definition?

"Dr. Claude De Ville, head of PAHO's Emergency Preparedness Program, pointed out that earthquakes have a profound health effect on the population. He urged citizens not to rush to bury the dead before identification, adding that 'while the presence of dead bodies is unpleasant, the cadavers do not cause disease.' Much more worrisome is the damage to the mental health of the survivors, who may suffer by not knowing if a loved one is missing or dead, he added." [PAHO, 16 Jan. 2001]

PAHO and WHO have been trying to overcome the myth that corpses are extreme health hazards for years and years. Why is this myth so persistent?

SUMA, the Humanitarian Supply Management System, Makes Novel Use of the Internet in El Salvador Earthquake

"El Salvador's National Emergency Committee (COEN) has activated the country's national SUMA team, whose members are among the more than 2,000 professionals trained in Latin America and the Caribbean. The team is setting up the SUMA system (www.disatser.info.desastres.net/SUMA) at anticipated points of entry of international aid to sort inventory and classify incoming humanitarian relief. At the request of El Salvador's government, PAHO and FUNDESUMA, the NGO that manages SUMA's logistical operations, sent a support team from Costa Rica to help in what is expected to be a major operation.

"The earthquake in El Salvador marks the first time SUMA has used the Internet to alert disaster-stricken countries about what is on the way. The Government of Colombia (whose national Red Cross Society helped to create the SUMA system and has been one of SUMA's strongest supporters in the Americas) has advised PAHO/WHO that they are using one of SUMA's specialized modules - the warehouse module - to register donations being collected by the Colombian Red Cross and Caracol, a local radio and TV station. Colombia will use the Internet to forward detailed information about their shipment to El Salvador's SUMA team, in advance of its actual arrival.

"Similarly, the National Emergency Commission in Honduras (COPECO) has activated its national SUMA team to register data on emergency supplies being collected at appointed locations, in coordination with the Red Cross and the Fire Department. As the supplies are en route to the neighboring country of El Salvador, Honduras' SUMA team also will have sent an advance report by Internet. This pattern of sending information on donations before the supplies actually arrive, using SUMA's standard software and criteria for classifying and assigning priorities to the supplies, will greatly aid the recipient country by allowing them to get the most important and urgently needed aid to those who need it quickly.

"FUNDESUMA is also mobilizing additional volunteers from the Dominican Republic, Venezuela, Honduras, Nicaragua, Colombia and Panama to support the team in El Salvador. The Governments of Honduras and Peru have also included SUMA trained  experts in their bilateral assistance to El Salvador." [PAHO, 16 January 2001]

There exist excellent technical people and systems in Latin America (recall Steve Bender's comment at FEMA focus group): engineers, geologists, geomorphologists, social scientists, public health and emergency physicians, etc., etc. Why can't this level of expertise be mobilized and effectively USED BEFORE a disaster, that is in prevention and mitigation - protection of schools, hospitals, critical lifelines, buildings? Both the building housing El Salvador's emergency commission and the offices of the El Salvadoran Red Cross were seriously damaged (recall observations of poorly constructed building where Red Cross delegation in Guatemala's second city, Quetzaltenango, was housed!).

"A 30-strong rescue team from Taiwan had the area sealed off and used sophisticated equipment to detect the slightest sound that would indicate a survivor." [APF, 17 Jan., 2001]

More technology.

Global Disaster Information Network GDIN Assistance to El Salvador:

"The Government of El Salvador requested assistance from GDIN-International www.gdin-international.org and within an hour, the organization began organizing assistance. At the request of GDIN, USG remotely sensed products should be posted up today. In addition, GDIN has requested help from the European Commission, the European and French Space Agencies, Spot Image and Space Imaging. ArcInfo has also provided assistance through GDIN. GDIN facilitated products come from many sources and will be posted on ReliefWeb. "

[Larry Roeder, Executive Director, GDIN, 15 January 2001, see http://www.state.gov/www/issues/relief/gdin.html ]

This is probably useful in damage assessment in a situation where transportation is difficult and there are many isolated rural hamlets. But, again, why is there is advanced technology applied AFTER the event, as with the medical inventory tool above, rather than as an aid to prevention? Does GDIN have access to images that would be of assistance in identifying landslide hazard? (See Landslide... is not Rocket Science.)

"450 people are confirmed dead in Santa Tecla. The town was founded in 1854 with the aim to build a new capital outside the zone of San Salvador's frequent earthquakes." [ACT, 16 January
2001]

Historical knowledge and historical irony! What about the one radio news mention of allegations that "rich" people were illegally harvesting trees from slope above Santa Tecla? Were the trees protected? Since when? Who was responsible for enforcement? Knowing vs. doing!!

Haresh Shah recalls his 1999 email discussion (hareshs@riskinc.com)

Thank you for your communications on El Salvador. I share your frustrations and feelings. After every event, whether it is in the developed world or developing world, we hear the same excuses and same expressions of "surprises". It seems to me that the societies at large have become very elastic. They keep taking in these excuses and "explanations" without breaking. Intellectuals keep talking, professionals keep meeting in conferences and workshops and what not, and the killing, misery, pain, disruption, keep happening with unfortunate regularity. What can we do? We cannot give up or get frustrated or point fingers. We must do what we can.

There are many, many ways in which we individually and collectively can help.

Besides the mailing list you have, I am also sending these communications to WSSI, people involved in US-Japan programs, a new alliance formed in Tsukuba, etc. I am also attaching herewith my communications with all of you after the Turkey, Greece, and Taiwan earthquakes. What I said almost a year ago still is valid. We all need to do something that will make a difference. Let us keep this dialogue going.

Landslide Hazard Identification isn't Rocket Science! Ben Wisner

...[R]escue crews reported pulling out four other victims from a truck swept away by a land slide along Guatemala's stretch of the Pan-American highway. Other people aboard the truck were listed as missing.

Authorities said 16 separate landslides had cut off roads around Guatemala and that at least 30 houses were destroyed or damaged in Jutiapa department when the earthquake hit at 11:33 am (1733 GMT) Saturday. [AFP, 14 Jan. 2001]

Landslides seem to be responsible for many deaths in El Salvador (200 homes buried in one location, 500 in another). Landslides following earthquakes are very common in Central America. Wouldn't it be relatively easy to identify areas along major transportation corridors (e.g. OAS trade corridors project) and also in towns and cities where hazard of landslide is high? Aren't there straight forward things that can be done to stabilize slopes, or, if not, then shouldn't people and critical facilities be relocated? This isn't rocket science.

"Landslide hazard mapping is not difficult and some has been done. We are trying to organize a project around the CA Pan American Highway." [Steve Bender, OAS, responding to Ben's mailings]

"Just outside the capital, workers dynamited massive hillside boulders to prevent them from crashing on the roads, several of which, including one major highway, reopened late Tuesday." [APF, 17 Jan. 2001]

Landslide hazard identification and mitigation, albeit crude. Why couldn't something of this sort have been done BEFORE? Among all the kinds of microzonation work that can be done BEFORE an earthquake, identification of zones subject to extreme landslide hazard must be among the easiest and most certain.

Notes on landslide hazard and risk identification and remediation 

by David Alexander, University of Massachusetts at Amherst 

To begin with, nothing reliable can be said about the landslides caused by the El Salvador earthquake until there is firm information on exactly what type of movement is involved (Varnes or other classification).

Secondly, landslide identification is, I venture to say, NOT a particularly difficult task. Most landslides occur in well-defined areas (e.g., stream headcuts). Many are reactivations of previous movements, or at least occur on slopes where one can detect signs of other movements in the past.

Pause for a brief anecdote: last July I went down to the regional air photo archive with the engineer from our local town hall. I showed him that under our town's middle school gymnasium there were signs of past landsliding (a hint of a degraded, curved scar on a slope undercut by a stream channel). Yesterday I heard that the brand new bypass around town will not open this week because immediately after being built it has become affected by landsliding. It passes under the middle school gymnasium. I have not yet been to see whether my prediction has been borne out (weariness and discouragement on my part, not lack of interest!).

The moral of this is that you can probably make quite a good prediction of where landsliding will occur by looking at reasonably sharp, black-and-white, panchromatic, stereographic aerial photographs printed at scales of 1:15,000 to 1:25,000, of the kind that any modern Zeiss repeating air photo camera will take (and, of course, you must make field visits to verify what you interpret). The cost of a sortie is usually about $10,000 per 150 sq. km, though it varies substantially (aircraft take-off costs are the main part). After that, you need nothing more sophisticated than the pairs of air photos ($10 - 15 each for 25x25 cm prints), a pair of stereo goggles ($19.95), a fine-point marker pen ($2.49), reasonably accurate contour maps (preferably at 1:10,000) and a few sheets of clear plastic. Of course, if you can also monitor some sites with piezometers and inclinometers, your results will be more accurate, but this probably will not add much to your diagnosis.

Landslide identification on aerial photographs requires (a) knowledge of the local lithology and what the terrain consists of and looks like, (b) consideration of texture, tonality and morphological signs that indicate potential landsliding, and (c) experience of the local and general scenarios of landsliding. Ceteris paribus, mottled terrain (signifying disturbed ground) and dark tones (signifying concentrations of groundwater) indicate a landslide hazard. Slope undercutting, oversteepening or incision, lithological junctions (e.g., clay abutting limestone at the surface), the presence of springs or seeps, and variations in vegetation (e.g. presence of hydrophytes such as canes, which colonize areas of high soil moisture) all indicate landsliding. Cuspate or lobate forms, headscarps, lateral shears, median trenches, and bulging ground, in either fresh or degraded form, are all indicative. The investigator needs to have in mind the characteristics of different types of movement (slides, flows, topples, falls, glides, Sackungen, rock avalanches; movements of rock blocks, sediment, earth, debris, mud; simple and composite types, etc.).

Other remote sensing methods of landslide identification are of dubious value. In general, landslides cannot be identified reliably on satellite images. Even colour air photos tend to be less effective than black and white ones. With the exception of liquefaction phenomena, most earthquake-induced landsliding is merely a matter of bringing forward in time, and possibly increasing in size and intensity, what would have happened in the absence of seismicity. It means that a whole suite of mass movements that would have occurred in any case (but separately) all happen together. Single earthquakes of M>7 can trigger 10 - 15,000 landslides, though not necessarily all at once, as some of them may occur within four days of the event as a delayed reaction because pore water pressure can take time to increase.

Earthquakes with heavy or persistent rain increase landslide potential. So do clastic sediments and thick soils on deeply incised terrain. The frequency of landslides increases with a roughly exponential function towards the epicenter, and the main concentration will occur within a radius of about 30 km.

Liquefaction is a special case, that is more closely associated with seismicity. Liquefaction potential can be predicted in advance of earthquake activity on the basis of knowledge of sediments (but in three dimensions, not just the surface pattern of deposits) and groundwater conditions. Lenses of sand within impermeable clays are particularly susceptible to liquefaction failure. So are clastic sediments arranged in alternating permeable and impermeable layers. The latter condition can give rise to lateral spreads-landsliding at very low angles - which is especially damaging if it carries along large (or giant) blocks of rock (e.g. olistoliths).

Slope stabilization is not a particularly mysterious process. If it is properly chosen and established, vegetation can bind soils together and reduce soil moisture content. But not all vegetation types are effective and some shallow-rooted trees can stimulate mass movement by blowing over in high winds (hence post-hurricane landsliding). Deforestation does increase the rate of landsliding, often by various orders of magnitude, but not automatically, because it also depends on other factors, such as what vegetation or land cover replaces the forest, what degree of weathering has taken place, what the slope drainage conditions are, whether the slope is steepening and lengthening, and how fast all these variables change.

Apart from revegetation and simple surface drainage works, most slope stabilization measures are expensive. Nets, flexible barriers, rock bolts, gunnite, terracing, deep drainage wells and channels, debris stilling basins and weirs, osmotic and cathodic electrical soil moisture reduction systems, pumps, excavation and regrading - they are all costly and can only be applied sparingly. At least one third of the cost will go in maintenance and operations - perhaps even two thirds.

Non-structural measures are cheaper and better than structural ones. It is easy enough to do a regional evaluation (nice if it is on a GIS) of factors That relate to landslide hazard (lithology, slope angles, lengths and orientations, vegetation and land cover, anthropogenic factors, etc). In the past it has taken me, working on my own, about a month to do this by hand for about 150 sq km, with no aids like GIS but with high accuracy. The end product is usually a regionalized map or rasterized matrix (perhaps with 1-hectare cells) of landslide potential categories from 0 to 4, which can be contoured if necessary.

Landslide risk requires - that the landslide potential map be crossed with maps of human habitation and land use. To do this meticulously is time-consuming and involves juggling with a variety of debatable assumptions about how vulnerability and hazard interact with each other, site by site. Nevertheless, it can be done quite effectively. Last year I did it for nine villages and towns in Umbria Region. The key is to look at past history of landsliding and use this to develop scenarios of what will happen in the future. In this particular field the past is truly the key to the present.

Further reading

Carrara, A., Cardinali, M., Guzzetti, F., and Reichenbach, P. 1995. GIS technology in mapping landslide hazard. In Carrara, A. and Guzzetti, F. (Editors), Geographical Information Systems in Assessing Natural Hazards.
Kluwer, Dordrecht: 135-175.

Carrara, A., Guzzetti, F., Cardinali, M. and Reichenbach, P. 1999. Use of GIS technology in the prediction and monitoring of landslide hazard. Natural Hazards 20(2-3): 117-135.

Cruden, D.M. and Varnes, D.J. 1996. Landslide types and processes. In Schuster, R.L. and Turner, A.K. (eds) Landslides: Investigation and Mitigation. Special Report, Transportation Research Board, National Academy
of Sciences, Washington, D.C.: 36-75.

Drennon, C.B. and Schleining, W.G., 1975. Landslide hazard mapping on a shoestring. Proceedings of the American Society of Civil Engineers, Journal of the Surveying and Mapping Division 101(SU1): 107-114.

Harp, E.L., Wilson, R.C. and Wieczorek, G.F., 1981. Landslides from the February 4, 1976, Guatemala earthquake. U.S. Geological Survey Professional Paper 1204A, 35 pp.

Lazzari, M. and Salvaneschi, P. 1999. Embedding a geographic information system in a decision support system for landslide hazard monitoring. Natural Hazards 20(2-3): 185-195.

Leroi, E. 1996. Landslide hazard-risk maps at different scales: objectives, tools and developments. In Senneset, R. (ed.) Landslides. Balkema, Rotterdam: 35-51.

Parise, M. and Jibson, R.W. 2000. A seismic landslide susceptibility rating of geologic units based on analysis of characteristics of landslides triggered by the 17 January, 1994 Northridge, California earthquake. Engineering Geology 58(3_4): 251_270.

Sidle, R.C., Pearce, A.J. and O'Loughlin, C.L. 1985. Hillslope Stability and Land Use. Water Resources Monograph Series, Vol. 11, American Geophysical Union, Washington, D.C., 140 pp.

Veder, C. 1981. Landslides and Their Stabilization. Springer-Verlag, New York, 247 pp.

David Alexander: Why don't we write an international standard on disaster mitigation and preparedness?