2012 Chapters (Layout Features)

International Research Institute for Climate and Society

Zubair, L.; Ralapanawe, V.; Yahiya, Z.; Tennakoon, U.; Chandimala, J.; Razick, S.; Lyon, B.

Although many natural disasters have hydro-meteorological antecedents, little advantage has been taken of the availability of weather and climate data, advanced diagnostics and seasonal predictions for disaster risk management. In this study, methodologies for use of hydro-meteorological data in hazard risk assessment are presented laying the ground work for future dynamic hazard predictions.

A high-resolution assessment of natural hazards, vulnerability to hazards and of multi- hazard disaster risk has been carried out for Sri Lanka. Drought, flood, cyclone and landslide hazards, and vulnerability were identified using data from Sri Lankan government agencies. Drought and flood prone areas were mapped using rainfall data that was gridded at a resolution of 10-km. Cyclone and landslide hazardousness were mapped based on long-term historical incidence data. Indices for regional industrial development, infrastructure development and agricultural production were estimated based on proxies. An assessment of regional food insecurity from the World Food Programme was used in the analysis. Records of emergency relief were used in estimating a spatial proxy for disaster risk. A multi-hazardousness map was developed for Sri Lanka. The hazardousness estimates for drought, floods, cyclones, landslides were weighted for their associated disaster risk with proxies for economic losses to provide a risk map or a hotspots map. Our principal findings are summarized below.

Useful hazard and vulnerability analysis can be carried out with the type of data that is available in-country. The hazardousness estimates for droughts, floods, cyclones and landslides show marked spatial variability. Vulnerability shows marked spatial variability as well. Thus, the resolution of analysis needs to match the resolution of spatial variations in relief, climate and other features. The higher resolution information is needed in planning and action for disaster management.
Multi-hazard analysis brought out regions of high risk in Sri Lanka such as the Kegalle and Ratnapura Districts in the South West and Ampara, Batticaloa, Trincomalee, Mullaitivu and Killinochchi districts in the North-East and the districts of Nuwara Eliya, Badulla, Ampara and Matale that contain some of the sharpest hill slopes of the central mountain massifs.

There is a distinct seasonality to risks posed by drought, floods, landslides and cyclones. Whereas the Eastern slopes regions have hotspots during the boreal fall and early winter, the Western slopes regions is risk prone in the summer and the early fall. Thus attention is warranted not only on Hot-Spots but also on “Hot-Seasons”.

Climate data was useful in estimating hazardousness in the case of droughts, floods and cyclones and for estimating flood and landslide risk. The methodologies presented here for hazard analysis of floods and droughts present an explicit link between climate and hazard. The results from this study coupled with the high-resolution seasonal climate prediction techniques developed in a related study point the way to using historical,
current and predictive climate information to inform disaster management policy, and early warning systems.

Climate, environmental and social change such as deforestation, urbanization and war affect the hazardousness and vulnerability. It is more difficult to quantify such changes rather than the baseline conditions.

Our analysis was carried out for a period since 1960 that included a period of civil war after 1983. This war affected the North-East of the island in particular. To put things in context, while natural disasters accounted for 1,483 fatalities in this period, the civil wars accounted for over 65,000. Wars and conflict poses complications for hazard and vulnerability analysis. Yet, the vulnerabilities created by the war make such efforts to reduce disaster risks all the more important.

Technical details of our work have been included in a case study published by the World Bank and in journals listed in the outputs.