An analytical Review of Selected Data Sets on Natural Disasters and Impacts

keywords:

Publication details

Publisher(s)	Centre for Research on the Epidemiology of Disasters (2006)
Author(s)	Liz Tschoegl

Description

In preparation for the ProVention Consortium’s Global Risk Identification Program (GRIP) Workshop on the Compilation of Reliable Data on Natural Disaster Occurrence and Impact, the United Nations Development Program (UNDP) and the Centre for Research on the Epidemiology of Disasters (CRED) have undertaken an analytical review of existing available historical disaster databases.

International databases: CRED; Munich RE: NatCat; Swiss RE: Sigma; ADRC: GLIDE; University of Richmond: Disaster Database Project; BASICS
Regional databases: La RED: DesInventar; ADRC
National databases: Australia: EMA Disaster Database; St. Lucia: St. Lucia Disaster Matrix; Canada: Canadian Disaster Database; United States: SHELDUS; United States: National Hazard Statistics; Philippines: DSWD-dromic; Philippines: NDCC Database
Sub-national databases: South Africa: MANDISA Database; UNDP/NSET: Nepal; NDP: Sri Lanka; UNDP: Orissa
Disaster events databases: Earthquake: USGS Database; Technological Disasters: UNEP/APELL; Flood: DFO Database; Flood: Hydrological Information Center; Tsunami: NGDC; Industrial Accident: MARS; Industrial Accident: MHIDAS
County-level events databases: Hurricane Mitch; El Salvador and Peru Earthquake; Indian Ocean Tsunami

Previous comparative analyses have recognized the following issues:

Fundamental differences in the definitions of disaster events and effects
continue to be problematic. A lack of standardization of the terminology used continues to complicate comparisons of data.
All of the databases continue to struggle with the issues of how to classify disaster types and sub-types and of reporting primary and secondary disasters. Without standardized terminology databases will continue to face a decreased precision in reporting the impacts associated with a disaster.
Georeferencing has allowed for more accurate recording of the location of disaster events but questions still remain on how to locate larger-scale disasters that cross borders as is the case with floods. While data resolution (by smallest administrative boundary) offers a detailed perspective not usually available, disaggregating the effects of a
disaster becomes difficult and may lead to overestimation of impact. Setting boundaries for disasters again is important if we wish to analyze the impact of similar events.
Issues of the level of resolution of a database are compounded by the difficulty of reporting the date of occurrence of a particular event. Events are recorded by some databases as falling within a range while others report a specific date, such as when a call for international assistance is made, which makes verifying same events difficult.
The lack of available detailed methodological information that is publicly available raises issues of the transparency of databases but also makes comparability difficult because of the ambiguity of important variables such as definitions, sources used, and inclusion criteria.
The availability of sources varies across the board. Whereas particularly developing country databases must rely on one source of information due to the lack of resources in the country, many developed countries struggle with trying to integrate and validate an overabundance of data sources.