Settlement Mapping and Morphological Analysis
The majority of modern urban planning in Africa is a direct product of colonial administration rooting back towards segregation and sanitation rather than integration. These outdated and Western models neglect the socio-cultural need of the local population, leading to unfulfilled potential for growth. [1]
Reports guiding basic concepts of urban planning in Africa, published by international organizations and agencies, emphasize "mesoscale" variables: zoning, density and regional connectivity, while neglecting "microscale" variables such as sidewalk quality, street lighting and social safety; variables that truly drive daily life in African cities. This disconnection leads to uncoordinated planning, traffic congestion, and haphazard sprawl. [2,3]
The binary categorization of "planned" (ordered) and "unplanned" (chaotic) is negligent and misleading of policy scales and targets: cities with top-down planning can exhibit higher entropy due to regulations targeted towards individual buildings rather than social spaces. [4]
Human societies construct a spatial culture by ordering their "spatial milieu": this organizes space to reproduce culturally specific patterns of social behavior. When measuring the level of order or disorder in cities, cities from different regions display distinct "signatures" of entropy. [1]
Western cities, such as Toronto, follow rigid, low-entropy grids while other cities follow organic, high-entropy orientations that may synergize with the local topography and social movement. If an idealized and "orderly" layout is imposed in these settings, the grid is the enactor destroying the innate natural and cultural identity. [2]
Informal settlements can be measured to provide the grounds for a formalized and functional planning system. Through terrestrial LiDAR scanning in the Rocinha region of Brazil, research discovered that informal settlements are not random, but follow consistent morphological rules based on the ratio of building height to street width as well as facade heterogeneity. [3]
The unconscious, yet culturally and anthropologically existent, decisions made in informal settlements create specific microclimates that regulate sunlight and airflow outperforming grids. These "street canyons" are especially critical for temperature control in hot climates. Instead of imposing a grid, planning can be a derivative of a set of "generative" codes that mimic the environmental performance of traditional settlements. [3]
Walkability as a concept does not simply refer to the existence of a road, but is governed by the surrounding ideas of accessibility, personal security, road safety, infrastructure quality and comfort. Mesoscale variables fail to account for these factors. [1]
The perception of walkability varies significantly by age and socio-economic status, with priorities ranging from infrastructure robustness to lighting; policy should be directed towards an inclusive nature and away from all-sized-fits-all standardizations. [1]
Nations recognizing mixed forms of settlements as planned, unplanned and informal, such as South Sudan, already have acknowledged the need for sustainable urbanization and land tenure security; in policy establishment and practice the integration of indigenous knowledge, departing from the Western-nominated norms, can help an organic, original and overarching movement for sustainable urbanization. [2]
Post (1996), writing on the politics of urban planning in Sudan, traces how the overall political climate, decentralization efforts, and prevailing administrative culture combine to produce planning systems that are chronically behind the facts of rapid urbanization. [1]
Juba has experienced among the fastest rates of urban growth in human history, growing from approximately 10,600 at Sudan's independence in 1956 to over 500,000 by 2012, a trajectory accelerated by the 2005 Comprehensive Peace Agreement and independence in 2011. Nearly 95% of the urban population resides in underdeveloped areas or informal settlements. [2]
Urban growth in South Sudan is driven primarily by conflict-induced displacement, the return of internally displaced persons and refugees, and the concentration of international development organizations in urban centers — what Grant and Thompson term the "Development Complex." [2]
// Unlike the industrialization-led urbanization of Western history, the spatial organization of South Sudanese cities is shaped by foreign investment and international organizations without adequate integration with local settlement patterns or governance.
Pareto's Southern Sudan Urban Development Strategy (2008) emphasizes that the lack of basic statistics and the turbulent pre-CPA period makes it extremely difficult to establish trends and estimate future urban growth. [3]
Martin and Mosel (2011) reveal that settlement patterns in Juba reflect a complex interplay of customary land systems, government land allocation, spontaneous occupation, and the spatial footprints of returning displaced populations. [4]
The Favelas 4D project by MIT's Senseable City Lab demonstrated that informal settlements follow consistent morphological rules quantifiable through terrestrial LiDAR scanning, proposing five key metrics: street width, street elevation, facade heterogeneity, facade density, and street canyon ratio. The methodology yielded 116 million points at 0.5-meter resolution. [1]
UN-Habitat shelter projects (2005–2012) in South Sudan constructed 8,300 shelters using two designs: bamboo and thatched-roof shelters for rapid deployment, and compressed mud block shelters with metal sheet roofs as a more permanent approach. Key lessons include the importance of community participation in design selection and the need to link shelter with basic services and livelihood programs. [2]
Netto et al. (2020) demonstrate that cities with rigid top-down planning can exhibit higher entropy than organic settlements, because regulations target individual buildings rather than the social spaces between them. The binary categorization of "planned" versus "unplanned" obscures the sophisticated spatial logic embedded in informal urbanism. [3]
Mayuol (2015) documents the politicization of public policies in South Sudan, where government decisions are often not implemented based on citizen expectations and where political interference undermines service delivery. Any planning framework must be robust enough to withstand political volatility and transparent enough to maintain legitimacy through community participation. [4]
The UN-Habitat publication Planning Urban Settlements in South Sudan (2012, updated 2020) compiles planning and development concepts for designing new settlement layouts and updating existing ones: grid-based planning, zoning classifications, lot and block dimensioning, and infrastructure planning for water, sanitation, roads, and electricity. [1]
This project does not propose discarding the UN-Habitat framework but extending it. The baseline infrastructure standards and institutional vocabulary remain valuable. What is needed is a complementary layer: a parametric, culturally-derived policy framework that reads the existing spatial culture and uses those patterns as generative codes for future planning. [1,2]
The National Urbanization Policy Framework consultative workshop (Juba, 2014) proposed harmonizing urban-regional standards across all states, but as Bandauko et al. (2020) document, the adoption of NUPs across the continent risks applying silver bullet solutions that jeopardize more suitable local policy innovations. [3,4]
Turok and Parnell (2009) argue that urbanization challenges in Africa are beyond the capacity of local government alone and require national-level policy coordination — but this coordination must be sensitive to the unique circumstances of each city and settlement. [5]
The methodology employs a tiered data acquisition strategy adapting the Favelas 4D framework to a data-scarce, post-conflict context where physical access to settlements is constrained by security concerns, limited road infrastructure, and geographic dispersion. [1]
The Global Human Settlement Layer (GHSL) from the European Commission's Joint Research Centre provides a baseline dataset for cross-referencing settlement identification with existing global databases. [2]
Machine learning classifiers — particularly convolutional neural networks (CNNs) — have been demonstrated to detect informal settlements with high accuracy. A ResNet-50 or U-Net architecture trained on labeled settlement data from comparable contexts (Nairobi, Dar es Salaam) can be transfer-learned for South Sudanese settlements, reducing the labeling burden. [3]
Acquire Landsat and Sentinel-2 time series for all ten state capitals and known secondary towns. Process in ENVI to generate multi-temporal land cover classifications. Cross-reference with OpenStreetMap data, UNHCR displacement tracking, IOM flow monitoring data, and the Global Human Settlement Layer.
Train a CNN classifier (ResNet-50 or U-Net) to identify and delineate informal settlement boundaries using labeled data from comparable East African contexts, with iterative refinement as South Sudan-specific training data becomes available. Use active learning strategies where the model identifies the most uncertain cases for human review.
Produce a national map of informal settlements with temporal layers showing formation, expansion, and relocation patterns from the 1980s to present. Overlay with geopolitical conflict data, natural disaster records, and displacement event timelines.
Select a stratified sample of settlements representing different regions, ethnic groups, settlement ages, and formation drivers: conflict displacement, economic migration, returnee resettlement, organic growth.
// OpenStreetMap data, analyzed topologically, has been used to detect informal settlements worldwide across 120 low and middle-income countries by analyzing street block accessibility patterns.
The Favelas 4D framework is built entirely around the "street scene" as its fundamental spatial unit — a street segment plus the building facades that directly face it. This presupposes a settlement organized around linear circulation corridors flanked by vertical facades: the morphological logic of a Brazilian favela like Rocinha. [1]
South Sudanese settlements are organized around a fundamentally different spatial logic: the compound. A compound is a cluster of structures — tukuls, shelters, fenced enclosures — arranged around a shared interior courtyard or open space, where the perimeter wall or fence, not the facade, defines the boundary between public and private.
The "global" analysis from Favelas 4D, which compares street scenes to one another, becomes a comparison of compound morphological profiles across the settlement. The "local" analysis, which subdivides each street scene into half-meter bands along the street's primary axis, becomes a radial or grid-based subdivision of the compound interior — measuring how metrics change from the compound center to its perimeter, or across sectors of the compound. [1]
// The RANSAC plane-extraction pipeline, designed to detect horizontal planes (streets) and vertical planes (facades), requires modification. South Sudanese structures often have curved walls (tukuls), sloped thatch roofs, and fence materials that produce noisy, non-planar point cloud signatures rather than the clean vertical planes of concrete and brick favela construction.
Each of the five Favelas 4D metrics must be rethought for compound-based typologies. Three of the original metrics — facade heterogeneity, facade density, and street canyon — are directly dependent on the presence and prominence of vertical facades. In South Sudanese settlements, where construction is overwhelmingly horizontal, these facade-centric metrics would return near-zero or undefined values — not because the morphology is uninteresting but because the metrics are not designed to capture it. [1]
// The informational richness in these settlements is in the horizontal plane: compound layout patterns, spatial relationships between structures, coverage ratios, perimeter configurations, and ground surface characteristics. The vertical bias of the original framework must be corrected.
Each compound receives a single morphological profile — a vector of compound footprint area, terrain elevation, structure height variance, structure density, and enclosure ratio. Compounds are compared to one another across the settlement, identifying clusters of morphological similarity and zones of divergence. Statistical analysis reveals whether consistent patterns correlate with ethnic groups, settlement ages, and geographic regions.
Compounds are subdivided into radial bands from centroid to perimeter, or into a grid, with metrics computed at each subdivision. This reveals internal spatial organization: whether structures cluster near the perimeter while open space concentrates at the center, whether construction height increases toward the back of the compound, how the transition from public to private space is spatially encoded.
The LiDAR data acquisition strategy inverts the Favelas 4D approach. Terrestrial scanning captures facades and streets well but misses roofs. For low-rise compound settlements, aerial LiDAR (drone-based) is the primary source — the morphological information is in the roof geometry, compound layout, and ground plane rather than in multi-story facades. Terrestrial scanning supplements with perimeter-level detail. [1]
Rocinha first emerged in 1927, with nearly a century of continuous development by the time of the 2020 LiDAR scans. The morphological complexity — variable facade heights, high heterogeneity, contrast between formalized courtyards and dense residential hillsides — reflects decades of incremental construction and vertical expansion. [1]
Young South Sudanese settlements — particularly displacement camps or newly formed returnee communities — would likely show significantly more morphological consistency. Early-stage settlements are built with uniform materials (UNHCR tarps, mud brick, thatch) at uniform scales (single-story, similar footprints). The within-settlement variation that Favelas 4D captures so effectively in Rocinha may be minimal in a settlement that is only months or a few years old.
Repeated compound-interior surveys show how the spatial organization of structures, open spaces, and perimeters evolves. The increasing availability of inexpensive LiDAR on mobile devices, which the Favelas 4D paper highlights as a scalability advantage, makes repeated temporal surveys feasible. [1]
While remote sensing provides the spatial morphology, understanding the human dimensions of settlement patterns requires additional data sources. In South Sudan, conventional human-centric data — census, household surveys, demographic records — is severely limited.
Call detail records from mobile operators have been used extensively for mobility analysis in other African contexts: tracking movement patterns, estimating population density, and identifying economic activity zones. CDR data can reveal the functional connections between settlements that physical infrastructure does not capture. [1]
Ma et al. (2021) scraped 874,588 traffic-related tweets in Nairobi, applied ML classification and geoparsing algorithms to create the first geolocated dataset of road traffic crashes for the city. In South Sudan, potential sources include geotagged posts and community-reported data through platforms like Ushahidi. [2]
Community mapping projects in Nairobi's informal settlements demonstrate how residents can be active participants in data collection. Organizations already operating in South Sudan — NRC, UNHCR, UNDP — have experience with community-based data collection that could be leveraged. [3]
The ultimate output is not a static planning document but a parametric policy framework — evidence-based design parameters that are adaptive, flexible, and culturally grounded. The compound-based metric framework produces the empirical basis for these parameters.
Instead of prescribing a 6-meter street width, specify an interstitial distance range demonstrated to maintain passage functionality and microclimate performance. Instead of a fixed setback, specify an enclosure ratio range derived from compounds that show the best environmental performance. [1]
// A parametric approach accommodates the diversity of settlement forms across different ethnic groups and regions. It allows incremental implementation — settlements can evolve within the parametric bounds without requiring wholesale redesign. The compound-based metrics make this possible in ways that street-based metrics cannot.
Develop generative algorithms that use compound-derived parametric codes to produce settlement layouts, enabling planners to visualize multiple culturally-consistent options for new development or settlement upgrading.
Translate the quantified compound morphological patterns (enclosure ratios, passage widths, structure density ranges, compound footprint areas, open space distribution) into computational parameters with defined ranges.
Use the parametric codes as inputs to generative design algorithms that produce multiple settlement layout options. Each iteration respects the compound-based cultural morphological bounds while optimizing for infrastructure access, microclimate performance, and emergency vehicle passage.
Layer culturally-derived compound parameters atop existing UN-Habitat infrastructure minimums for water access, sanitation, road widths, and fire separation distances. The parametric layer adds the morphological ratios, connectivity patterns, and spatial organization principles that existing frameworks lack.
Present generated layouts to communities for feedback. As the temporal metrics (consolidation rate, densification rate, formalization index) accumulate data, the parametric ranges are refined. The framework is a living document — continuously updated as settlements are re-surveyed and environmental conditions change.
// In the longer term, the compound morphological data and parametric models could form the basis of digital twins for South Sudanese settlements — computational models that simulate the effects of proposed interventions before implementation.
Extending the South Sudan framework into the broader morphological memory methodology requires decomposing settlements into discrete spatial units — morphological cells — and computing the influence exerted between them by geography, infrastructure, policy, and resistance. [1]
Fleischmann et al. (2020) formalize the concept of the Morphological Cell (MC) through Voronoi tessellation of building footprints, producing a plot-proxy entity that captures spatial properties at the finest meaningful scale of urban form. These cells allow systematic measurement of dimension, shape, spatial distribution, intensity, connectivity, and diversity between adjacent units. [2]
Hillier's Space Syntax framework identifies a dual structure in all cities: a foreground network of longer, straighter connections linking centers at all scales, and a background network of shorter, right-angled, grid-like local structures. The foreground network tends to converge across cultures; the background network is culturally idiosyncratic — configured to restrain and structure movement in the image of a particular spatial culture. [3]
// The morphological memory of a settlement is encoded in the background network. When a foreign spatial system (highway, grid, policy regime) is imposed, the foreground network is overwritten while the background network either resists, adapts, or deteriorates.
By separating cells along temporal, infrastructural, and policy boundaries, the project can assess which cells were most consequential in shaping surrounding form, which were most influential in propagating a spatial logic, and which experienced the harshest morphological deterioration — losing connectivity, density, or cultural pattern coherence. [1,2]