MONDAY DEVELOPMENTS The Latest Issues and Trends in International Development and Humanitarian Assistance
The Great Balancing Act Conservation & Human Development
April 3, 2006 Published by InterAction Vol. 24, No. 6
Knowledge of Physical Processes Can Assist Natural Disaster Response By M.H. Bulmer
T
he year 2005 witnessed some of the worst natural disasters in modern times. The extent of physical damage and numbers of people killed, injured or displaced varied as did the relief and development responses. What is common to each of these disasters is that they were the result of natural processes that operate at a range of scales in both time and space and are an intrinsic part of the evolution of landscapes. For example, globally in 2005, there were eleven earthquakes of magnitude 7.0 or greater, but only three occurred in densely populated areas. Each of these events leaves behind a physical record that it occurred, which if not altered or erased, can be used to understand how the landscape has evolved.
underpinned by a scientific understanding of the natural hazards and the risks associated with them.
The number of serious natural disasters reported in 2005 has raised questions about our understanding of the recurrence interval of natural disasters, whether they are becoming more frequent, whether human activity is partly responsible, and what should be our response. In the case of landslide disasters, a range of contemporary processes such as deforestation, population growth, migration and climate change are likely to continue to lead to increasing impacts, more than offsetting the current rate of improvements in understanding and mitigation.
In recent disasters, scientists have been actively discouraged from traveling to disaster zones for fear that their work will interfere with the humanitarian efforts on the ground. There has been little understanding at the senior management level of aid agencies of the importance of obtaining scientific data to determine how the physical event can be derived, or that there may be additional events. Critical information about a natural disaster is often short lived and easily altered or obliterated by subsequent natural processes or human activity. For example, rains in the days after the 1999 Chi Chi earthquake in Taiwan altered the geomorphic signature of the earthquake to the extent that the original landslide inventory was lost. The need to arrive at disaster affected areas within days to collect data on these short lived signatures is critical.
The need for answers is becoming acute as aid agencies and lending organizations struggle to meet the rising costs of natural disasters. Reducing the costs of natural disasters through mitigation strategies has become the stated policy of organizations such as the Inter-American Development Bank because the costs of reconstruction are unsustainable. However, mitigation can only be effective if it is
10
MONDAY DEVELOPMENTS
Currently, the major science funding agencies in the United States are not mandated to conduct mitigation research. This means that scientific research is only conducted once a disaster occurs, funded through emergency grants. However, 2005 revealed some significant weaknesses with such a system because of the number of disasters and the time that they occurred. Landslide disasters in Guatemala, for example, occurred within the same timeframe as Hurricane Katrina, which took funding preference. Clearly, there is a need to re-examine whether this mechanism is working in the way that it was intended.
Our experience reveals that disaster response and reconstruction agencies do not routinely collect the types of data required to understand the physical processes that
caused the natural disaster. For example, the majority of the 20,000 deaths associated with Hurricane Mitch (1999), the 30,000 deaths in Venezuela (1999), and the 4,500-plus deaths in Haiti/Dominican Republic (2004) were the result of debris flows triggered by slope failures rather than the floods as often reported in the media. Such knowledge can be of real value to relief and reconstruction efforts and can be obtained through increased collaborations between relief agencies and scientists.
loss of life, if this issue is not addressed. Relocation off the fan will significantly reduce the risk.
In recent disasters, scientists have been actively discouraged from traveling to disaster zones for fear that their work will interfere with the humanitarian efforts on the ground. There has been little understanding at the senior management level of aid agencies of the importance of obtaining scientific data to determine how the physical event can be derived, or that there may be additional events.
The magnitude 7.6 Kashmir earthquake that struck on October 8, 2005, as a result of a rupture on Muzaffarabad Fault, provides a study of how science and knowledge of the landscape can assist relief, development and reconstruction agencies. The official death toll was 87,300, with landslides responsible for about 20,000 fatalities. A reconnaissance visit by our team in January 2006 revealed that although large numbers of landslides occurred, the total is substantially less than would be expected for an earthquake of this size. It is clear, however, that many other slopes have sustained high levels of strain. Some of the slopes are currently moving, and a few have failed since the earthquake. It is likely that in most cases failure has not yet developed because of the exceptionally dry conditions that have prevailed prior to and since the earthquake, which has led to low pore pressures. But once the monsoon rainfall arrives, due in July 2006, many of these slopes are likely to fail.
During our visit, it became apparent that some refugee camps are situated in highly dangerous positions. In several locations, particularly in the Balakot area of Northern Pakistan, tented villages are located on alluvial fans, which are sand and gravels deposited by a stream where it emerges from a narrow valley, that are young and fresh, suggesting high levels of activity prior to the earthquake. The slopes in the catchments show signs of multiple shallow rock slope failures triggered by the earthquake. In the event of heavy rain, further failures are expected to occur and the sediment from the existing failures will start to move again. These are near-perfect conditions for the generation of debris flows and flash floods. Clearly, the resistance capacity of tented villages to such events is effectively nil. Thus, we anticipate the potential for large-scale
In a number of other locations, we noted that tented villages have been constructed on low river terraces. These locations are threatened by normal rises in river level during the rainy season, exacerbated greatly by the potential for larger than normal floods associated with impounding and then breaching of landslide dams. Again, we see a very high potential for disastrous outcomes if this problem is not addressed in the near future.
In the medium to long term, two possible scenarios in the earthquake affected area may occur: firstly that the slopes will steadily stabilize and revegetate; and secondly that the area of landsliding increases significantly, stripping the slopes of vegetation. Our observations in January indicate that the second scenario is more likely and that the 2005 earthquake has allowed for the development of very large, active landslide systems that will remobilize during every large rainfall event. The impact of these landslides on the local community and infrastructure are likely to be severe and river valleys may become blocked. The consequence of an increase in landslide occurrence will be a marked increase in the production of sediment from the slopes. Downstream of the landslides will show increasing suspended sediment concentration through time. At present, there remains a large volume of sediment on the slope that is now disturbed but has not yet reached the fluvial system. The downstream consequences could be severe. Clearly, there are many unknowns in the earthquake affected area in Pakistan, but the signs on the ground point to major risks of near-term landslides, with significant implications for ongoing and planned relief, reconstruction and development efforts. M.H. Bulmer is Director of the Landslide Observatory in JCET at the University of Maryland Baltimore County. For more information visit, www.landslidemitgationgroup.org. Email questions and comments to
[email protected]. Photos: courtesy of M.H. Bulmer.
APRIL 3, 2006
11
