Y1 Progress report | Agnes Kapinga
Community-based approaches for nature-based solutions in flood-prone areas: The case of the Dar es Salaam Metropolitan Development Project, Tanzania
Fieldwork in the flood-prone wards of Dar es Salaam offered both encouraging and challenging experiences. Ward leaders and youth were consistently willing to guide the research team to specific sites and households once the purpose of the research had been explained, and were generally forthcoming, often raising concerns about flooding that went beyond the scope of the questionnaire. Building this level of trust nonetheless required repeated visits rather than a single introduction, particularly in areas where earlier research or development interventions had left commitments unfulfilled.
Engagement with communities also revealed how flood risk in these areas is rooted in a longer history of land-use change. Many affected areas were originally agricultural land, converted to residential use as residents chose to remain in the city following the privatisation of state industries rather than relocate to rural areas. The resulting informal settlements continue to attract migrants drawn by proximity to the city centre and to employment opportunities. Livelihood pressures, in turn, sustain practices that exacerbate flooding: sand mining along riverbanks persists despite its recognised contribution to flood risk, often destroying riverine trees planted specifically for flood mitigation, while individual households frequently construct informal barriers to keep floodwater out of their homes, inadvertently obstructing drainage and worsening flooding for neighbours.
Livelihood pressures, in turn, sustain practices that exacerbate flooding: sand mining along riverbanks persists despite its recognised contribution to flood risk, often destroying riverine trees planted specifically for flood mitigation, while individual households frequently construct informal barriers to keep floodwater out of their homes, inadvertently obstructing drainage and worsening flooding for neighbours. Viewed together, these dynamics resemble the behaviour of a complex adaptive system rather than a set of independent problems.
Viewed together, these dynamics resemble the behaviour of a complex adaptive system rather than a set of independent problems. Individual households and sand miners act as autonomous agents responding rationally to livelihood needs, yet their combined, decentralised decisions produce flood outcomes at a scale none of them intended or could control individually. The interaction between land-use conversion, migration, livelihood-driven sand mining, and household-level flood barriers forms feedback loops in which adaptations to flood risk at one level (an individual wall, a mining site) increase exposure at another (a neighbouring household, the wider drainage system). This suggests that flooding in the study area is best understood not as a fixed hazard to be mitigated through isolated technical interventions, but as an emergent property of interacting social, economic, and environmental agents, one that will continue to adapt and shift in response to any single intervention introduced into the system.
Phase One (Q1–Q2) of the first year centred on two stakeholder workshops, one on the development of data collection tools and one on validation of findings, which shaped the methodological approach applied in subsequent fieldwork. These workshops were supported by ground-truthing of nature-based solutions across five wards in Dar es Salaam and the collection of GPS ground data for remote sensing and GIS analysis. Sixteen key informants were interviewed during this phase. Findings from the validation workshop were disseminated through a public lecture at the Tengeru Institute of Community Development, extending their reach beyond the immediate project team. This phase was accompanied by attendance at a scientific writing workshop in Senegal.
Phase Two (Q3–Q4) built on this foundation through participatory geospatial data collection and mapping of the spatial distribution and community-perceived effectiveness of nature-based solutions, with analysis of this dataset now in progress. A household questionnaire survey was concurrently initiated, with 220 households surveyed to date across two of the five study wards; administration will continue into the next phase to cover the remaining three wards. This phase was accompanied by engagement with CDKN Ghana on science communication and participation in a co-creation workshop in Uganda on climate change communication tools.
Stakeholder engagement was sustained throughout the year. The Dar es Salaam City Council Environmental Officer served as the focal point, coordinating local research activities. CDKN Ghana provided technical assistance on science communication. Ward leaders supported community mobilisation and administrative coordination. Local community members participated through the household survey and geospatial data collection, contributing local knowledge that informed both data quality and interpretation. Moreover, the academic staff of the Tengeru Institute of Community Development provided academic input. With Phase One complete and Phase Two data collection ongoing, the upcoming quota will prioritise completing the household survey in the remaining three wards and finalising the geospatial analysis, while also making progress on the manuscript currently under review.
At the start of this project, I understood transdisciplinary research primarily as the combination of academic disciplines, in this case, GIS and remote sensing, social science, and urban planning, brought together within a single research design. This understanding has shifted toward a broader view that places non-academic actors and their knowledge on equal footing with academic methods, rather than treating community and stakeholder input as supplementary validation of technical findings.
This shift became apparent during Phase Two, in the ground-truthing and stakeholder workshops, where the interpretation of remote sensing data depended heavily on local knowledge held by ward leaders and community members, knowledge that could not have been generated solely through GIS analysis. The workshop on co-creating climate change communication tools in Ghana further reinforced this, as it required translating technical findings into formats that could be jointly shaped and understood by policymakers, practitioners, and community representatives, rather than disseminated to them after the fact.
The science communication workshop also influenced this shift, showing that transdisciplinary work extends beyond data collection into how findings are framed, packaged, and made usable for actors outside academia. Administering the household questionnaire survey in Phase Two deepened this further, as it became clear that the 220 households surveyed were not simply data sources but participants whose own framing of flooding and nature-based solutions shaped how the research questions themselves were being refined. By the end of the year, academic and non-academic knowledge were no longer separate inputs; I combined them at the analysis stage. They were shaping each other throughout the research, from how questions were framed to how findings were communicated back to the people who had contributed to them.
One peer-reviewed article was published in City and Environment Interactions (Elsevier) and was based on ground-truthing work conducted in Dar es Salaam. A workshop report on the co-creation of a science communication toolkit, developed through engagement with CDKN Ghana, has been published. A stakeholder engagement report covering the Dar es Salaam study area was also produced, documenting engagement with local stakeholders across the reporting period.
No major challenges were encountered over the course of the year. A minor delay occurred in Q2 due to time spent on workshop attendance, tool refinement, and household survey administration, which required careful sequencing around community availability and access in informal settlement areas. However, this did not affect the overall project timeline. Trust-building with community members was an ongoing consideration throughout data collection, addressed through participatory approaches, adherence to community entry protocols, and continuous engagement with ward leaders, who supported access to survey sites.
Annual report submitted by Dr Agnes Kapinga
(summarised for publication by Leti Kleyn for the FAR-LeaF Programme)
