Home | About Us | Contacts | Help | Feedback | SEARCH Global | Australia | NSW | Vic. | Qld | WA | SA | Tas. | ACT | NT | Ant. | Weather & Warnings | Hydrology | Climate | Numerical Prediction | About Services | Learn About Meteorology | Registered User Services |

Global Guide
Main Menu

Global Guide to Tropical Cyclone Forecasting:
CHAPTER 7: WARNING STRATEGIES

7.5 HAZARD, VULNERABILITY AND RISK ASSESSMENT

As discussed in Section 7.3, because of the inherent forecast uncertainty, tropical cyclones initially offer a macroscale threat to large areas and significant populations. This condenses to a mesoscale highly destructive impact in which the greatest loss of life and property generally occurs in small areas comprising a small subgroup of the autonomous local government authorities initially under threat. Since the warning task is to persuasively alert and to promote response to counter this developing threat, it follows that any developments in the warning system should commence with hazard risk and vulnerability assessments of the elemental local government authority areas. When integrated, these indicate the total vulnerability of all components of the system, districts, provinces, and nations. However, each house has its own vulnerability and only sustained community programs can hope to bridge the gap between the formality of government protection at the higher level and the street -by-street level of the local community.

In several advanced countries, notably the USA, detailed hazard risk and vulnerability assessments, often initiated by meteorologists themselves, have been performed over much of their cyclone-prone coastal zones for many years. This preparedness homework has enabled detailed evacuation planning to be implemented on the basis of operational processing of the SPLASH and SLOSH storm surge models in accord with bathymetric and topographic risk mapping, assisted by community behavioural response studies in the most vulnerable sectors. This capability is incorporated in recommended response activity announced in hurricane local statements issued by national weather service offices around the US coast that follow authoritative national hurricane advisories issued by the NHC. Detailed topographic mapping makes it possible to recommend evacuation on a street-by-street basis in respect of forecast storm surge inundations augmented by heavy rainfall runoff.

Regrettably it is recognised that although most other countries have published information on the frequency of cyclone occurrences, and some have computed return periods for cyclones of stated intensity or cyclone parameter statistics, only a few countries have undertaken comprehensive vulnerability assessments, especially at community or district level. The absence of adequate hazard mapping and skills may have contributed to this situation. A result is the almost complete absence of localised contingency action plans in these countries, which negates many of the benefits of improved warnings and inhibits the deployment of warning strategies.

7.5.1 Quantification of Risk

Vulnerability and risk analysis provides a structured analytical procedure to identify and quantify hazards and to estimate the probability and consequences of their occurrence. It must be emphasised that the absolute risk is a complex, multiplicative function of the hazard level and the vulnerability of a community. In an illustrative sense, this means that:

Disaster Risk = Hazard x Vulnerability.

Say we were to rate the hazard and vulnerability on scales of 1-5, then a community with high vulnerability and hazard levels of 5 would be many times more at risk (25) than would a community with low levels of 1.

7.5.2 Hazard Assessment

Although many countries use designated cyclone warning phases or stages, few have adopted scales which combine a rating of the intensity of cyclone parameters with corresponding estimates of the typical damage that may be expected. Examples of this categorisation include the US Saffir-Simpson scale, and the Australian Cyclone Severity scale (Section 8.5), both of which range from 1-5. It is recommended that such categorisations be undertaken by all countries with a tropical cyclone threat.

The hazard component may then be assessed in terms of the frequency return period of a particular intensity category for the specified location. This method is particularly useful in countries where detailed observations are not available, and assessment is best made using derived parameters from the known tropical cyclone statistics. Alternatively, return period estimates of particular cyclone parameters, such as maximum wind gusts, storm surge heights, river flood levels or rainfall intensity may be developed or estimated. These latter criteria are favoured for engineered structural mitigation measures. In any case an estimate is required of the return period frequency of cyclones of each intensity category as guidance to provide an objective indication of the cyclone hazard in each coastal sector.

For the purpose of general community preparedness a worst-case scenario for a period of 30 years may be reasonable. If this is translated into appropriate contingency planning, any cyclone of lesser intensity should not cause major unexpected adverse effects (provided community vulnerability does not increase). Preservation of essential community lifeline facilities (hospitals, power and water supplies, communication systems, meteorological radar tracking stations, cyclone shelters and operations centres etc) require design criteria for a longer time period.

National standards associations generally determine regional engineering design wind loadings. An indicative example, normalised by Australian Cyclone Severity Scale (or category), for a stretch of coast from Port Douglas to Fraser Island on the Australian East Coast and applicable to 50 km inland is (SAA, 1989):

These statistics refer to cyclones directly impacting coastal areas. Approximately twice as many cyclones in each category threaten the coastal section, without direct impact.

An indication of return periods for tropical cyclone parameters for Andhra Pradesh and West Bengal (which provides a guide for Bangladesh) has been published by Jayanthi and Sen Sarma (1988) based on a 95-year data base. This includes the following statistics for maximum wind speed and storm surge height:

The two cyclones that occurred in nearly the same locations in Andhra Pradesh in 1977 and 1990 and the two that occurred in Bangladesh in 1970 and 1991 were of a similar intensity that lay within the 25-50 year return period in this table. This illustrates the problem with literal use of return period statistics. Although catastrophic, with 300,000 and 140,000 deaths, respectively, in Bangladesh, and 10,000 and 1000 deaths in Andhra Pradesh, these statistics illustrate the advances in warning-response systems in these regions.

7.5.3. Vulnerability Assessment

Vulnerability comprises the people in their environment and their exposure to the cyclone hazard. The geographical and physical components of provincial, district and local level vulnerability are assessed through a combination of hazard risk maps, detailed demographic maps, and information on supporting community infrastructure and facilities. Such assessment can be made by satellite remote sensing, aircraft photography, and on the ground by surveyors, engineers and urban and rural planners. A good survey will provide details of:

• Siting and constructional integrity of community lifeline facilities, including access roads;
  
  
• Proportions of flimsy, partly cyclone-resistant, and cyclone resistant residential housing, commercial buildings and markets;
  
  
• Existence and state of maintenance of cyclone protective works, river and coastal dykes and embankments, drainage systems and public shelters;
  
  
• Degree of protective forestation and mangroves, which can reduce wind, wave and surge energy.

The non-physical social and economic aspects of vulnerability of a community also need to be assessed. These include a measure of the community's capacity for coping with the occurrence of cyclones by martialling resources and organising effective response actions. Readily assessable elements include:

• The existence of a local counter-disaster council, an equipped emergency operations centre, a cyclone contingency plan of action which has been resourced, implemented and rehearsed;
  
  
• Evidence of an ongoing community cyclone awareness program.
  
  

An obvious indication of the vulnerability of a community may be found from its performance in coping with a previous cyclone, after noting subsequent improvements.

An extensive program of interdisciplinary training courses entitled "Improving Cyclone Warning Response and Mitigation" is now conducted by the Asian Disaster Preparedness Center (ADPC), based at the Asian Institute of Technology in Bangkok. The ADPC has introduced practical experience for participants through the conduct of field vulnerability assessment studies in towns and villages. Such studies take advantage of any available hazard maps and primarily comprise visual inspections of potentially hazardous terrain, community life and facilities, aided by interviews with officials and residents on cyclone and flood hazard experiences and on the current state of preparedness measures, local communications and warning arrangements. The course participants, comprising meteorologists and hydrologists, disaster managers, engineers, planners and technical personnel interacting in workshop sessions. Reports to a workshop plenary session, supported by maps and sketches, are discussed and published and the total exercise takes little more than a day's work.

While these exercises obviously comprise a fair degree of subjectivity they summarise a good deal of human hazard experience of residents in a short time and probably acquire about 80% of the environmental, physical, and socio-economic information needed for a useful vulnerability assessment. An immediate feel for the community's hazard awareness is gained, as well as knowledge of the cyclone-resistant integrity and maintenance standards of the principal lifeline facilities.

A recent finding (PAHO, 1992) notes that a nation's effectiveness in disaster prevention, mitigation and preparedness is related to its level of economic activity. This indicates that current socio-economic indices, such as per-capita income, may provide one objective basis for quantifying vulnerability in the above equation.

A complementary input to the vulnerability index should come from recent disaster experience. For example, damage evidence in relation to 20 categories of socio-economic activity is clearly noted for the April 1991 Bangladesh cyclone in BCAS (1992). This cyclone effected 10,800,000 people with a total damage of US$2.1 billion, or roughly the per-capita GDP of US$210 (FEER, 1993) This simple analysis could be extended to local districts and normalised on a scale of 1-5. Separate quantification could be developed for deaths and casualties.

Where recent damage statistics are not available, the current per-capita income probably provides an objective basis for relating statistics from other regions and quantifying both the economic and social effects of a tropical cyclone.

7.5.4 Potential Disaster Risk Scale

No objective method has yet been devised to integrate the cyclone hazard and vulnerability into a disaster risk scale similar to the Saffir-Simpson Intensity scale. Such a scale would first require the development of a suitable vulnerability scale, as recommended in the previous section, then a method needs to be developed for incorporating this with the hazard scale. It is almost certain that the total risk will be a multiplicative combination of the hazard and vulnerability scales, but research is needed to determine the optimum combination.

Such a scale could markedly simplify the development of warning and response strategies and the allocation of mitigation measures. It could provide a globally consistent indication of disaster risk. The scale could be printed on maps and displayed on computer workstations as indicators of the relative vulnerability of hundreds of autonomous local government areas threatened by an approaching cyclone. National potential disaster risk maps can be prepared and disseminated to indicate an objective measure of the real danger of the threat, compared to previous occurrences, and thus to promote preparedness measures. By agreement an tropical cyclone warning centre could include such information in its advisory warning messages, together with potential disaster risk predictions for each community, in a manner reminiscent of the issue of hurricane strike probabilities in the USA.

Contents Chapter 7.6