Abstract
The deliverable introduces a map of the vulnerability to flooding in Dar es Salaam at the resolution of the finest administrative level (the subward/mtaa; comprising approx. 5-15 000 residents). Overlaying a hydrological model, representing the areas of the city most likely to become flooded, the high-risk areas may be identified. That is, where the flood-prone areas coincide with the highly vulnerable subwards.
The deliverable explores in a step-by-step manner how to capture, measure and process spatial data of multiple dimensions and integrating them into a Geographical Information System (GIS).
The overall approach is a spatial multiple criteria evaluation (S-MCE) process, following a series of steps, whereby the most important multi-dimensional indicators of vulnerability to flooding are selected, measured and analyzed. Eventually the output is presented as a product in one, aggregated dimension, easy to comprehend for policy- and decision-makers.
The deliverable also presents how the stakeholders of a city are introduced to this process at an early stage, and how they as an expertise group, are providing with vital information and insight, that give the study relevance and are facilitating the subsequent steps of the methodology. The preceding Deliverable 3.3 compiled a comprehensive list of relevant indicators of vulnerability to flooding. This deliverable (D3.4) takes off from there with the selection and weighting of indicators by the stakeholders. Subsequently the stakeholders provide valuable input to (proxy) variables to the indicators and how they may be measured. An important conclusion from these participatory processes is that the selected set of most important vulnerability indicators is highly distinct in space (site-specific). The stakeholder interactions in Dar es Salaam and Addis Ababa show that the selected indicators are only partially overlapping, which would discourage from generating a common subset of indicators for all flood-prone cities.
Taking the standpoint in the multi-dimensional setup (a conceptual framework developed by CLUVA Task 2.3), the deliverable presents ample examples of how to work with indicators spatially. Detailed introductions are given to how the more GIS-related indicators, like mobility and low-lying areas, as well as the more intangible indicators, like institutional capacity and trust, may be measured and mapped.
However, taking on the task of an S-MCE approach in a developing country also reveals its limitations. The methodology here was developed with the comprehension that the data availability and accessibility might be limited. Consequently, there was a hope that the bulk of the input data would be already available at project start – generated from conventional GIS procedures and originated from routine data collection techniques. However, later in the project it turned out that some parts of the data supply was insufficient, and that alternative sources had to be consulted, which initiated a lengthy and difficult process of data accessibility.
The final vulnerability map for Dar es Salaam is indicating that vulnerability is strongly associated with the informal residential areas (characteristically the unplanned settlements), but the relationship is not fully conclusive. Thus, there are examples of densely populated low-income residential areas where the vulnerability not necessarily is high.
Taking the flood-prone areas into account shows that the main high-risk areas (subwards) are located to the west of the city center. Yet again, looking at the individual vulnerability indicators, they reveal that there is no common subset of indicators explaining the vulnerability of the subwards in the high-risk areas. Consequently, the indicators relating to high vulnerability of one subward in the high-risk area, may not be the same in a nearby subward in the same flood-prone area.
Furthermore, there is a distinction between the vulnerable subwards in the peripheral areas compared to them closer to the city center. Among the subwards in the outskirts of the city the vulnerability is to a greater extent associated with a few of the Asset dimension indicators (e.g. Income and Age). Closer to the city center the vulnerable subwards are more linked to the indicators of the Physical dimension (e.g. Low-lying Areas, Population Density, Dangerous Infrastructure/Industry). These patterns may possibly be related to the attraction of the city center and the job opportunities found there. The greater job opportunities in the city center make it favourable to settle there. This produces a scarcity of available land to dwell on and generates a high population density. As a result, people with fewer resources are staying in the crowded settlements closer to the center, or are taking the risk of settling on the nearby flood-prone floodplains. The opposite alternative for people with fewer resources is to settle in the less dense peripheral areas and away from the low-lying areas, but there the job opportunities are fewer.
Interestingly, our results are indicating that in the sense any formal and more affluent communities are expressing vulnerability to flooding, the vulnerability is explained by the indicators of the Institutional and Attitudinal dimensions (e.g. Participatory Decision-making, Level of Social Network).
The deliverable explores in a step-by-step manner how to capture, measure and process spatial data of multiple dimensions and integrating them into a Geographical Information System (GIS).
The overall approach is a spatial multiple criteria evaluation (S-MCE) process, following a series of steps, whereby the most important multi-dimensional indicators of vulnerability to flooding are selected, measured and analyzed. Eventually the output is presented as a product in one, aggregated dimension, easy to comprehend for policy- and decision-makers.
The deliverable also presents how the stakeholders of a city are introduced to this process at an early stage, and how they as an expertise group, are providing with vital information and insight, that give the study relevance and are facilitating the subsequent steps of the methodology. The preceding Deliverable 3.3 compiled a comprehensive list of relevant indicators of vulnerability to flooding. This deliverable (D3.4) takes off from there with the selection and weighting of indicators by the stakeholders. Subsequently the stakeholders provide valuable input to (proxy) variables to the indicators and how they may be measured. An important conclusion from these participatory processes is that the selected set of most important vulnerability indicators is highly distinct in space (site-specific). The stakeholder interactions in Dar es Salaam and Addis Ababa show that the selected indicators are only partially overlapping, which would discourage from generating a common subset of indicators for all flood-prone cities.
Taking the standpoint in the multi-dimensional setup (a conceptual framework developed by CLUVA Task 2.3), the deliverable presents ample examples of how to work with indicators spatially. Detailed introductions are given to how the more GIS-related indicators, like mobility and low-lying areas, as well as the more intangible indicators, like institutional capacity and trust, may be measured and mapped.
However, taking on the task of an S-MCE approach in a developing country also reveals its limitations. The methodology here was developed with the comprehension that the data availability and accessibility might be limited. Consequently, there was a hope that the bulk of the input data would be already available at project start – generated from conventional GIS procedures and originated from routine data collection techniques. However, later in the project it turned out that some parts of the data supply was insufficient, and that alternative sources had to be consulted, which initiated a lengthy and difficult process of data accessibility.
The final vulnerability map for Dar es Salaam is indicating that vulnerability is strongly associated with the informal residential areas (characteristically the unplanned settlements), but the relationship is not fully conclusive. Thus, there are examples of densely populated low-income residential areas where the vulnerability not necessarily is high.
Taking the flood-prone areas into account shows that the main high-risk areas (subwards) are located to the west of the city center. Yet again, looking at the individual vulnerability indicators, they reveal that there is no common subset of indicators explaining the vulnerability of the subwards in the high-risk areas. Consequently, the indicators relating to high vulnerability of one subward in the high-risk area, may not be the same in a nearby subward in the same flood-prone area.
Furthermore, there is a distinction between the vulnerable subwards in the peripheral areas compared to them closer to the city center. Among the subwards in the outskirts of the city the vulnerability is to a greater extent associated with a few of the Asset dimension indicators (e.g. Income and Age). Closer to the city center the vulnerable subwards are more linked to the indicators of the Physical dimension (e.g. Low-lying Areas, Population Density, Dangerous Infrastructure/Industry). These patterns may possibly be related to the attraction of the city center and the job opportunities found there. The greater job opportunities in the city center make it favourable to settle there. This produces a scarcity of available land to dwell on and generates a high population density. As a result, people with fewer resources are staying in the crowded settlements closer to the center, or are taking the risk of settling on the nearby flood-prone floodplains. The opposite alternative for people with fewer resources is to settle in the less dense peripheral areas and away from the low-lying areas, but there the job opportunities are fewer.
Interestingly, our results are indicating that in the sense any formal and more affluent communities are expressing vulnerability to flooding, the vulnerability is explained by the indicators of the Institutional and Attitudinal dimensions (e.g. Participatory Decision-making, Level of Social Network).
| Original language | English |
|---|
| Publisher | University of Copenhagen |
|---|---|
| Number of pages | 133 |
| Commissioning body | EU 7th Framework Programme |
| Publication status | Published - 2014 |