Topic 5.2: Economic Analysis

Rapporteur: R. J. Murnane
Risk Prediction Initiative
Bermuda Biological Station for Research, Inc.
PO Box 405
Garrett Park, MD 20896, USA

Email: rmurnane@bbsr.edu
Fax: 301.942.1886

Working Group: R. Falls, J. Jarrell, M. Yumoto


Human suffering and economic losses are the common result of tropical cyclone landfall. The Economic Analysis sub-group of the Tropical Cyclone Impacts group examined a number of issues related to economic losses from tropical cyclone landfall as one step in a long-term effort to ameliorate the effects of tropical cyclone landfall. After reviewing the status of past recommendations from the IWTC-IV this sub-group report examines a number of steps that can be taken to improve our ability to quantify and assess past losses from tropical cyclone landfall and help mitigate future suffering and loss. We encourage: 1) the development of an international data base on natural disaster losses, costs, and mitigation expenses; 2) support for enhanced best track data sets for all ocean basins subject to tropical cyclones; and 3) the construction of a parametric wind field model that accounts for recent advances in our understanding of tropical cyclone wind fields.

5.2.1: Introduction

Tropical cyclones cause massive suffering and huge economic losses. The Working Group on Tropical Cyclone Impacts at IWTC-IV examined the societal and economic impacts associated with tropical cyclones and how warnings can help minimize future losses. This section will review the suggestions from the Economic Impact sub-group, assess any progress over the past four years, and suggest future actions that would minimize economic impacts and maximize planning and preparedness for future tropical cyclones.

Economic costs associated with natural disasters are defined as losses that are reimbursed through insurance or by governments . Thus, true economic losses are often significantly larger than economic costs, particularly for regions where people or businesses do not commonly use insurance.

Estimates of economic losses from natural disasters are rare and are derived mainly from insurance records. Government estimates of economic costs and/or losses are often derived as a multiple of the insured cost estimates. A number of global reinsurance companies provide via the world-wide web (e.g., <http://www.swissre.com>) summaries of insured costs and deaths from catastrophes. It is worth noting that the insured cost estimates are often adjusted for inflation. However, the costs often do not account for changes in population density and wealth . We will use the term "cost normalization" to describe costs that have been corrected, or normalized, to account for changes in population, wealth, and inflation that occurred since the event occurred.

Losses from natural disasters and tropical cyclones can be classified as either direct or indirect . Direct losses are associated with the direct effects such as physical destruction. Indirect losses are associated with the consequences of the destruction, e.g., unemployment and business interruption. Indirect losses are difficult to quantify if they are not costs covered by insurance.

Losses can also be subdivided into "avoidable" and "unavoidable" losses. Avoidable losses are those that can be avoided through mitigation measures, including well-integrated warning and response systems. Unavoidable losses occur when mitigation action cannot be taken to counter the effects of a potential hazard, or cannot reduce the resulting loss. Changes in knowledge and approaches will alter through time what is considered unavoidable (e.g., Handmer et al. 2002).

Accurate estimates of economic losses from tropical cyclones and other natural disasters would help emergency management, planners, governments, and businesses plan and cope with tropical cyclones and other natural disasters. A data base on the costs of disaster mitigation efforts are required in addition to a record of losses to properly assess the costs and benefits of efforts to mitigate damage from tropical cyclones and natural disasters.

5.2.2: Review and Status of Past Recommendations from the Economic Impacts Sub-group

  1. Trends and Methodology

Countries were encouraged to undertake studies that would normalize to present-day values their historic costs and losses from tropical cyclones. There have been a few such studies within the United States (e.g., , we are not aware of such studies within other countries.

In Queensland Australia, various Government and University supported projects (the Tropical Cyclone Coastal Impacts Program (TCCIP), AGSO Cities Project, Queensland Local Government Disaster Mitigation Project, and the Queensland Greenhouse Task Force projects) have produced detailed studies of risk from storm surges, riverine flooding and wind impacts (as well as geological hazards) for a number of coastal cities at the household to community level. Few of these studies have included economic assessments; however, they provide data that could be utilized or adapted for that purpose. In addition, there have been exploratory studies of the economic impact of tropical cyclones on the tourism industry (Drake 1995; Sofield 1993), and on community losses for insurance purposes (e.g., Oliver and Georgiou 1994); however, much remains to be done.

In Australia, a Natural Disaster Risk Management Studies Program supported by the three levels of Government has been launched. A manual Disaster Loss Assessment Guidelines (Handmer et al. 2002) has recently been issued, with a companion worked example.

b) International Data Base on Losses and Costs of Disasters

The working group recommended the development of a global damage database of damage statistics from the start of the century. The data would include negative as well as positive impacts and be collected in a common format. This work would probably require the full-time effort of a number of experts and include efforts from other disciplines such as economics. There has been little formalized progress towards this goal. However, the National Research Institute for Earth Science and Disaster Prevention (NIED) of Japan recently started a new project titled "Study on extreme weather events and water-related disasters due to Climatic Change". One aspect of this effort is the development of a new database system about typhoon activity, disasters, and social impacts by typhoon in Japan.

c) Literature Review

A final recommendation of the group was the establishment of a working group of experts to review the international and national literature on the economic and social aspects of tropical cyclones worldwide. The working group would use the historical data to predict future economic and social losses and costs and assess the ability of tropical cyclone warning and information systems to reduce damage. This recommendation has not been acted on.

5.2.3: International Data Base on Losses and Costs of Natural Disasters and Expenses of Mitigation

It is important to gauge the economic value of tropical cyclone warning and information systems. A cost-benefit analysis will need accurate statistics on the costs and losses of past events as well as the expenses for tropical cyclone warning and information services and for mitigation efforts. The losses and costs of past events will need to be normalized to account for changes in population, wealth, and inflation.

We encourage the WMO to promote the development of an international data base that would track the losses and costs of tropical cyclones as well as the expenses associated with tropical cyclone warning and information services and mitigation efforts.

5.2.4: Enhanced Tropical Cyclone Best-track Data Sets

The Warning Process sub-group of the Tropical Cyclone Impact Group at IWTC-IV recommended the development of improved and enhanced best-track data. We reiterate this recommendation because of the value of best-track data for a number of uses, including their role in economic analysis. The relevance of best-track data to economic analysis is of particular relevance to risk models that estimate the probability of future losses and costs from tropical cyclones. In addition, risk models are capable of producing normalized data bases for losses and costs from natural disasters and for predicting the probabilities and costs of future events.

Risk models are made of three basic components: a hazard component, a damage component, and a loss component. The hazard component provides the probabilities and characteristics of a specific hazard. For a tropical cyclone, it would provide information such as storm track, forward motion, wind field, and decay over land. The damage component of the risk model uses this information to calculate the expected damage and the loss component estimates costs associated with rebuilding, loss of business, and the liability of a specific company.

Best-track data are used to define probability distributions for characteristics of tropical cyclones (e.g., forward motion and intensity) in a risk model. Additional storm characteristics such as the radius of maximum winds, maximum gusts, and rainfall would improve the depiction of tropical cyclones by the hazard component of a risk model and lead to more realistic estimates of future losses and costs and improve estimates of the benefits of mitigation efforts.

At the 34th Session of the ESCAP/WMO Typhoon Committee a small Working Group was formed to produce a detailed proposal for the development of a unified Northwestern Pacific Tropical Cyclone Best-Track Data Set. The plan would be submitted to the Typhoon Committee at their 35th Session for their consideration. We encourage the extension of this effort to all regions subject to tropical cyclones.

5.2.5: Parametric Wind Field Model

Parametric wind models form the basis of the tropical cyclone hazard component of many risk models. A next-generation parametric wind model would provide a more realistic wind field and improve the tropical cyclone hazard component of a risk model and thereby improve loss and cost estimates associated with tropical cyclone landfall. Recent wind field observations collected with Global Positioning System (GPS) dropsondes have many characteristics that are not represented by standard parametric models.

The Analysis sub-group of the Tropical Cyclone Prediction Group from IWTC-IV encouraged the development and testing of parametric models. We agree with this and also encourage the development and testing of next-generation parametric wind models.

5.2.6: Summary

We review the status of economic issues associated with tropical cyclones that were identified at IWTC-IV and find at best modest progress. We then discuss several topics of economic relevance and encourage the following steps to stimulate further progress:

1) The development of an international data base on the economic losses and costs associated with tropical cyclone landfall. The data base would include information that would allow the losses and costs to be normalized for changes in population and wealth and inflation in the affected areas. To maximize the usefulness of the data for cost-benefit analysis, the data should include the expenses associated with tropical cyclone warning and information services and mitigation efforts.

2) The enhancement of tropical cyclone best-track data sets in all ocean basins and regions affected by tropical cyclones. The data sets should include additional parameters beyond the standard information on location, maximum sustained winds, and central pressure. The additional parameters could include the radius of maximum winds, maximum gusts, rainfall, time and location of landfall, and other additional information over sea and land. The enhanced best-track data set would improve the hazard component of tropical cyclone risk models, improve the performance of statistical intensity forecast models, and be helpful for emergency management and planning.

3) The development of a next-generation parametric wind model. This model would account for recent wind observations generated by GPS dropsondes. Among other benefits, the improved parametric wind field would lead to more realistic loss and cost estimates produced using tropical cyclone risk models and improve the initial wind fields for numerical simulations used for forecasts and research.


Anaman, K.A., Lellyett, S.C., Drake, L., Noar, P.F., Sullivan, P.J., Henderson-Sellers, A., Thampapillai, D.J., 1997: Economic and Social Benefits of Meteorological Services provided by the Australian Bureau of Meteorology. Macquarie University/Bureau of Meteorology Final Project Report.

World Meteorological Organization, 1994: "Meteorological and Hydrological Risk Management", WMO/TD No 591, TCP-32.

Session 5.2