Issues around developing and working with vulnerability indicators at the urban level in Europe

Submitted by Daniel Morchain | published 17th Apr 2012 | last updated 21st Sep 2015

Image: City of Meissen flooded by the River Elbe, 2002 (Marc Zebisch)

Vulnerability indicators can be used to help estimate measures or trends of possible future harm by measuring past or present vulnerability. As such, they are a valuable tool to identify (qualitatively or quantitatively) vulnerable spots, guide the allocation of adaptation funding, communicate and raise awareness, and monitor the effectiveness of policies. Nevertheless, the difficulty of expressing vulnerability (in this context understood as it has been defined by the IPCC) and how its three components: exposure, sensitivity and adaptive capacity should be combined, together with other shortcomings, such as insufficient data, make the development and analysis of vulnerability indicators as a measure of vulnerability only a partial exercise in visualising the whole picture.

This article discusses European experiences with vulnerability indicators at the urban level, as well as difficulties, advantages and recommendations for working with them. It is based on the scoping study by Schauser, I. et al  titled “Urban Regions: Vulnerabilities, Vulnerability Assessments by Indicators and Adaptation Options for Climate Change Impacts”, published by the European Environment Agency and the European Topic Center on Air and Climate Change in December 2010.

There are numerous examples of using vulnerability indicators to build vulnerability assessments. The purpose of any given assessment will determine the type and objective of the indicators that will need to be produced, and their temporal and spatial scale. It will also help to inform the more appropriate level of aggregation of the indicators. 

A vulnerability assessment should clearly define its boundaries in order to produce relevant and usable results. Boundaries should include the definition of the system/area being considered (the potentially vulnerable system/area), the type of climate stimulus that it may be threatened by (just one per indicator to avoid confusion in determining causality between climate stimuli), the type of damage that may be expected, and the time scale. Stakeholder involvement in the development of vulnerability assessments will prove valuable in developing a comprehensive view of the area assessed, and will increase the acceptance of results and of potential adaptation measures to follow.  

In Europe, indicators of the three components of vulnerability have been mostly developed in relation to heat waves and fluvial floods, with fewer available for urban flood, sea level rise, water scarcity and wild fires. These mostly reflect vulnerability in the past or present because of the difficulty, uncertainties and lack of data associated with future adaptive capacity (on the other hand,future exposure and sensitivity values may generally be obtained through climatic projections and population dynamics). 

However, inserting adaptive capacity components in a meaningful way is crucial for assessing vulnerability, though it is difficult to accomplish - even when looking at past conditions - due to lacking specific data and the impossibility or high cost of obtaining it. Alternatively, generic information is often used (GDP per capita, literacy levels, percentage  of the total population made up of elderly citizens, etc.), but its appropriateness is open to question. Expert knowledge and views, in any case, are frequently used in the development and aggregation/weighting of indicators, although the theoretical background for their development is often missing. Data availability always plays a crucial role in determining their selection.

When considering city-wide vulnerability, it is important to not limit the boundaries strictly to the urban region, as cities depend on their surroundings and other areas of influence (not necessarily geographically close) for their sound operation. These lifelines include water, food and energy supply. 

Notwithstanding the effort needed to construct indicators, it is often difficult to (re)use them outside of the scope and specific location for which they have been originally created. This is because (i) the specific data used to build the indicator may not be available in other regions, and (ii) the aggregation methods of indicators are frequently complex and not always made public. This can be considered a drawback of vulnerability indicators, as their scope for use and their potential for comparison and benchmarking becomes limited. However, succumbing to the temptation of applying them across regions without proper adaptation can deliver inaccurate results - for instance, populations in warmer regions may respond differently to temperature increases than people in colder regions. Still, some authors believe that transferability should be one important criterion of a good vulnerability indicator. 

As the figure below illustrates (p. 45 of the scoping study), if the purpose of the assessment is general (high-scale, coarse), then standard quantitative indicators can be used to produce a rough understanding as well as a comparison e.g. between countries or regions. These assessments are normally conducted with a top-down approach. Increased resource and data availability at a local scale (city, village) can produce more context-relevant quantitative as well as qualitative data resulting in more comprehensive and localised assessments (usually conducted with a bottom-up approach).

To conclude, it would be worthwhile to consider the following knowledge gaps at the European level (at least some are already being addressed): 

  • The development of a method for selecting and aggregating vulnerability indicators is lacking. Any method to be developed should be “simple, easily reproducible and transparent for the user. (...) Weighting of variables should only be done based on or at least verified by stakeholder and expert opinion.”
  • The availability, comparability, quality and management of data at levels beyond the local needs to be improved. Presently lack of data is a major obstacle for the development of vulnerability indicators.
  • More knowledge is needed on how vulnerability indicators may be further used to support adaptation efforts.
  • There are only few projects evaluating the implementation of vulnerability indicators (as opposed to their development).  
  • Conceptual links between vulnerability indicators and frameworks/tools to work with policies need to be improved.