2.1 Methodological framework
To identify the sites with highest potential for sand dams in the province of Namibe, we use a mixed-methods approach building on quantitative and qualitative methods applied to ten communities in the two municĂpios of Virei and Bibala. This approach was specifically developed to meet two major challenges in the siting of water harvesting in drylands. First, drylands, especially in Sub-Saharan Africa, are often areas with poor data availability and difficult access to information (Prinz et al., 1998). Lack of high resolution spatial biophysical data is a major limitation to estimating a complex water harvesting technology as sand dams. A suitable river section for sand dams construction must have specific characteristics such as presence of high rocky riverbanks, adequate size of the stream, shallow bedrock, coarse and little clayish sediments and low salinity (Beswetherick et al., 2018; Ngugi et al., 2020). Although some of these data might be available, they usually come from global-scale datasets, rarely tested in Sub-Saharan Africa and at a coarse resolution, which could compromise the reliability of the analysis. The second challenge is represented by the complete lack of socio-economic data. Although with some uncertainty, the literature on best-siting of water harvesting structures suggests the use of socio-economic parameters as distance to communities, rivers, roads, land tenure, population density, workforce availability or prior community experience with similar infrastructures. However, these kinds of data are rarely available and particularly lacking in Angola, where most of scientific and technical work is outdated and definitely neglected remote areas such as rural Namibe (Amado et al., 2020).