Annotated Bibliography

by Yuetian (Peak) Xu

1. Hiraishi, Tetsuya (2000). Tsunami Risk and Its Reduction: Integration of Disaster Management Technologies. Proceedings of 3^rd EQTAP Workshop. Manila, Philippines

This paper discusses an initial study of tsunamis in the Asia Pacific region and comes up with a list of recommendations at the end. Of particular interest is the result of the study. They found that in the 2000 Luwuk-Banggai tsunami in Indonesia, there was water run up of 2-3m approximately 10 minutes after the earthquake. This sets a very tight timeframe for a sufficient warning system. A simulation with installing a belt of mangroves appears to be effective at preventing damage from the water pressure generated from the first wave. This article is interesting and applicable as it reports the conditions a solution must meet as well as a solution that met with qualified success.

2. Okusai, H. (2000) Tsunami Disasters and the Countermeasures in Japan, 20th Century. [WWW Document] URL

http://www.tsunami.civil.tohoku.ac.jp/hokusai2/topics/counter.html (visited 2005, September 21).

This is a list of Tsunamis in Japan during the past century and how countermeasures worked or did not work on each occasion. In general, before the Tsunami Forecasting System was started by the Japan Meterological Agency, the casualties were about an order of magnitude higher and in the thousands rather than hundreds. By the late 1960’s most sea walls were already in place and the system was working well. This meant that the warnings were getting through and death tolls were in the double digits with many of them from the earthquake rather than the tsunami. However, in May 1983 and July 1993 tsunamis, there were critical warning delays of 13-15 minutes and 5 minutes respectively which cause a significant increase in death toll. The official conclusion was that the tsunami waves got around the sensors. This emphasizes the density of sensor distribution as integral in our final solution.

3. Hawkes, Nigel (2004, December 29). How earthquake jolted the planet. Times, p.14

This is a news item that puts in layman’s terms the power of the earthquake. This is useful in appreciating the strength of what we are dealing with. Magnitude 9 earthquake is roughly 190 million tons of TNT. The tsunami uses that force as well as that from underwater topology shifts to form its wave. The resulting impact of the tsunami in December 2004 was able to shift the coastlines of Sumatra by up to 120 ft. This raw power brings into doubt the ability of dissipation systems like sea walls and levees despite their successes in Japan. This may mean a careful examination of what the underlying situation is like in each case and perhaps a tiered solution to target areas with varying magnitudes of tsunamis.

4. Japan Times (2004, December 29). Mexico disaster center holds first tsunami prevention confab with Japan. Japan Times, p.23

This is another news article. This time, it’s discussing international cooperation. The meeting of experts from both sides of the Pacific represents a step forward. Japan is donating $12 million for the establishment of a tsunami center and monitoring system as well as prevention program. This shows how little such a system can cost. A comment that was interesting was the Japanese motivation to help Mexico. They claimed they wanted to help since the Mexico City earthquake of 1985 killed 10,000 people. It’s terrible that it takes a tragedy of such magnitude for a developed country to want to help prevent another.

5. IEEE Computer Graphics and Applications (2005). Near Real-Time Tsunami Computer Simulations within Reach. IEEE Computer Graphics and Applications v. 25 no. 5 (September/October 2005) p. 16-21

This article discusses the issue of computer simulation of tsunamis. Within roughly four hours of the quake, there were detailed and accurate simulations generated that modeled the impact of the tsunami in Indian Ocean and beyond. The problem was that is impressive as it may sound, it’s too little too late. The author points out that with current technology, one can cut this down to the order of minutes and almost real time. This will help with later disasters such as the waves that hit Mexico but is still insufficient for neighboring areas. If our final solution involves a simulation, it must be better than real time. Given realistic warning delays and the time constraints, any simulation would have to be generated in something close to 5 minutes.

6. Heller, Valentin; Unger, Jens; Hager, Willi (2005). Tsunami Run-up--A Hydraulic Perspective. Journal of Hydraulic Engineering v. 131 no. 9 (September 2005) p. 743-7

This group in Switzerland discusses the tsunami run up. They point out that from their point of view, it’s difficult to stop such a quantity of water. The more realistic alternative would be diversion. This appears to be another vote against sea walls and the like. Their discussion of the mechanics of wave hydraulics made me think of attempting to create destructive interference between waves but details are fuzzy as of the moment.

7. Johnson, Frank (2005). Working Toward an Indian Ocean Tsunami Warning System. Sea Technology v. 46 no. 8 (August 2005) p. 19-20, 23-4, 26.

This is a discussion of the current system being built in the Indian Ocean by the UNESCO. This is the first article to mention satellite technology as a monitoring tool. The idea of geosynchronous satellites combined with in situ sensors makes for more accurate measurements. There is also some mention of more need for global communication in the warning network.

8. Weinstein, Stuart A.; Okal, Emile A. (2005). The Mantle Magnitude Mm and the Slowness Parameter J: Five Years of Real-Time Use in the Context of Tsunami Warning. Bulletin of the Seismological Society of America v. 95 no. 3 (June 2005) p. 779-99.

This paper discusses two different measures and how they studied them. The first, Mm, was introduced by in 1989 as is a measure of the magnitude. This was found to be biased due to an excess of monitoring stations found in North America. This is compensated for by a weighted calculation. The second part mentions a measure to determine size of tsunami earthquake combination and to differentiate aftershocks from resulting tsunami. They mention using a sliding window which is a machine learning idea that I used before in computer science research. Simple recording of tremendous amounts of seismic data and training upon it will produce reasonably high accuracy rates for quick detection. A principled system can also aid in discovering patterns difficult for humans to discover.

9. Beasley, Craig J. (2005). Connecting with other societies: A report from Southeast Asia with a tsunami update. Geophysics v. 24 no. 6 (June 2005 supp) p. 568, 574.

This is an article exploring the cultural barrier for researchers. The author sent to the scene of the December 2004 tsunami discusses the impact of the tsunami and also how he communicated with the locals. This brings up the issue that certain engineering solutions may not be acceptable to the local populace. In addition, education programs need a distinct cultural awareness component.

10. Powers E. Michael (2005) ASCE Study Finds Seawalls Were Effective in Tsunami. ENR v. 254 no. 17 (May 2 2005) p. 20.

The ASCE engineers were positioned at five beaches in Sri Lanka, India, and Thailand. They assessed that sea walls were not sufficient to prevent damage by themselves as the destruction was still significant. However, they claimed that up to 60% of the horizontal momentum was deflected vertically. Designs should consider sea walls to be merely a first line of defense as well as changing the structure to deflect more horizontal momentum. Perhaps differing angles should be in place for differing points. Too much attempted deflection in the first line of sea walls may simply cause structural failure.

11. Dengler, L. A. and Magoon, O. T. (2005). The 1964 Tsunami in Crescent City, California: A 40-Year Retrospective. Proceedings of Solutions to Coastal Disasters 2005

This article discusses the 1964 tsunami which caused great devastation. Nevertheless, they mentioned the tsunami awareness brought about by this. The United States then set up the tsunami warning system in concert with Japan which prevented any major tsunami damage to Japan for most of the 60s and the 70s and the United States since then.

12. Ichii, K. and Donahue, M. J. (2005). Evaluation of Sea Dike Settlement Due to Seismic Shaking Prior to Tsunami Attack. Proceedings of Solutions to Coastal Disasters 2005.

This study showed that many sea dikes, even those made of concrete can settle up to 2 meters in a severe tsunami. This severely compromises their operation. They recommend additional reinforcements as a counter. The main problem is not just a lower dike but a structurally less sound one. This is quite interesting and leads to an idea to perhaps add some spring/flotation based system to maintain the dike upright at all times.

13. Walters, Roy A. (2005). Numerical Simulation of Tsunami Generation, Propagation, and Runup. Proceedings of Estuarine and Coastal Modeling 2005.

This paper discusses a pressure based tsunami model for New Zealand. This model fits well with the data from the December 2004 tsunami. A key appears to be the study of pressure gradients on a macroscopic scale. This model, with slight modifications, should work well for Micronesia. It might not work as well for Peru due to some presumptions upon sea floor topology unique to the Western Pacific.

14. Erdman, Craig, Preuss, Jane, Barnett, Elson T., and Murphy, Vivyan (2003). Planning and Mitigating for Local Tsunami Effects. Proceedings of Advancing Mitigation Technologies and Disaster Response for Lifeline Systems 2005.

This is a documentation of tsunami damage to structures based on Crescent City tsunami of 1964. The findings are that much of the damage was done to the structures via apertures such as windows where the water rushes in. In addition, the back wash after a wave causes a negative pressure gradient that destroyed already weakened structures. The lesson is that merely making the front side of a building resistant is insufficient. The overall design should have tsunamis in mind. Furthermore, the buildings on the shoreline may experience up to a magnitude more force than those just half a mile inland. This point to ample possible work in zoning.

15. Borrero, Jose C. (2002). Field Survey of the June 23, 2001 Earthquake and Tsunami in Southern Peru. Proceedings of Solutions to Coastal Disasters ’02

This tsunami in Peru hammers home several points. The waves were actually strongest away from the strongest point of quake. The water can reach up to 1km inland, even in Peru with its high elevation gradient. The water receded for 15-20 minutes before rushing back up, giving ample time for people with instinct/education to clear the area. Many areas were absolutely swept clean by the tsunami. Buildings with strong foundations and concrete have highest survival rates meanwhile pressure gradients caused walls of adobe buildings to blow out. Many bamboo, adobe buildings were completely destroyed. A lightweight, cheap, and durable building material is needed.

16. Yoon, Sung B.; Choi, Chul-Sun; and Yi, Sung-Myeon (2002). Numerical Modeling of Tsunami Propagation over Varying Water Depth. Proceedings of Ocean Wave Measurement and Analysis.

This model works better than others by virtue of accounting for the propagation speed variations due to water depth. They had different ways to handle the water depths between 1/25 and ½ the tsunami wave length. Their model’s output corresponded well with the data from several recent tsunamis. A direct result of their model is the fact that a section of the sea floor with wildly varying water depth is very effective at absorbing tsunami energy. Perhaps a way to artificially change depth of a small section of high risk coastline might be in order.

17. Wood, Nathan; Good, James; and Goodwin, Robert (2002). Reducing Vulnerability of Ports and Harbors to Earthquake and Tsunami Hazards. Proceedings of Solutions to Coastal Disasters ’02

Harbors would be utterly devastated if there is an earthquake or tsunami near the mouth of the harbor or port. The primary aim is to prevent a far off tsunami from creating excessive damage. The main concern is what is known as harbor resonance. This is a very destructive build up of succession of waves. This can be countered by barrier structures outside and better interior design as to create destructive interference. This article is particularly useful as it addresses a very important set of economic structures to preserve and also a set of high risk, low amount of coastline to protect. There can be custom solutions tailored for harbor protection due to this different type of problem. Micronesia does not appear to have any major harbors but this may come in handy for Peru.

18. Bernard, Eddie N. (2002). The U.S. National Tsunami Hazard Mitigation Program. Proceedings of Solutions to Coastal Disasters ’02

This discusses the current sensor and warning system in place for the United States Pacific area. This can give some insights into how to adapt the current system for our purposes. They mentioned that the US program can inform the entire chain of command about tsunamis within an hour or two. This is still due to the low density of sensors and some need for more streamlined process. This led to non-intercepted cases such as the two tsunamis in Japan mentioned earlier. This means that though the system that exists is far better than anything elsewhere, it’s still grossly inadequate.

19. Hamzah, M. A.; Mase, Hajime; and Takayama, Tomotsuka(2000). Simulation and Experiment of Hydrodynamic Pressure on a Tsunami Barrier. Proceedings of Coastal Engineering 2000

This is another study upon a simulation of forces upon a tsunami barrier. This fits well with previous tsunamis yet did not show the amount of deflection of horizontal momentum that was observed by on site engineers in the December 2004 tsunami. This shows, to some extent, the fact that models can at times fail and one should not overly rely on them. Thus, perhaps rather than a model, a simple learning/recognition system would work better.

20. M. González and R. Medina(1998). Probabilistic Model for Tsunami-Wave Elevation Along the Alborán Seacoast. Proceedings of Coastal Engineering 1998

This is yet another model. What is interesting is that it includes a distribution of tsunami wave heights. This is very helpful practically as the height of seawalls constructed cost a lot of money. With an idea of the probability of tsunami waves exceeding a particular height and then causing damage, we can run a thorough cost benefit analysis to determine the optimal height of the seawall. This should come with certain qualifications as the model is not necessarily perfect. Nevertheless, with such a model, this makes seawalls a far more practical solution.