Background
Methods
Study area and data
General framework
Hazard–a/Larval habitat suitability
Identifying a set of criteria (factors and constraints), and obtaining or producing the corresponding geospatial layers
Land-cover classes from VHR imagery | Alternative existing open product(s) | Larval habitat suitability—Sub-Saharan African cities | Larval habitat suitability—Dakar |
---|---|---|---|
Buildings | Open buildings | Buildings are not among sites likely to provide artificial breeding sites [9] | Buildings do not provide suitable habitat conditions. There are very few water bodies on flat roofs and balconies [21] |
Swimming pools | n/a | Swimming pools should not be discarded as potential breeding sites [57] | |
Paved surface | OSM | Paved surfaces (e.g., roads and parking lots) are not cited among common artificial urban breeding sites [9] | Paved surface is unlikely to provide suitable conditions for the occurrence of breeding sites [21] |
Dumpsites | OSM | Breeding sites tend to be located close to human dwellings rather than in uninhabited areas such as the large city dumpsite [28] | |
Bare ground | Esri 2020 Land Cover, WorldCover | Temporary water bodies (e.g., puddles) can form on bare ground in the wet season [21] | |
Grass | Esri 2020 Land Cover, WorldCover | ||
Scrub/shrub | Esri 2020 Land Cover, WorldCover | Scrub/shrub is not among features likely to provide natural breeding sites [9] | Breeding sites tend to be located close to human dwellings rather than in uninhabited land, such as areas covered in scrub/shrub [28] |
Trees | Esri 2020 Land Cover, WorldCover | ||
Small water bodies < 100 m2 | n/a | Small water bodies are more likely to host larvae than medium-sized and large water bodies [28] | |
Medium water bodies 100 to 1000 m2 | OSM, Esri 2020 Land Cover, WorldCover | Medium-sized water bodies can also be utilized as breeding sites [9] | |
Large water bodies > 1000 m2 | OSM, Esri 2020 Land Cover, WorldCover | Large water bodies can also be utilized as breeding sites [9] | |
Water courses | OSM, Esri 2020 Land Cover, WorldCover | ||
Marine waters | n/a | Marine waters are not favourable to An. gambiae. An. melas is found in low numbers in such waters, but it is a less efficient malaria vector [34] | |
Shadow | SRTM-derived shadow, using the GRASS GIS module r.sunmask |
Land-use classes from VHR imagery | Alternative existing open product(s) | Larval habitat suitability—Sub-Saharan African cities | Larval habitat suitability—Dakar |
---|---|---|---|
High-density planned residential areas | WUDAPT LCZ | In densely built-up areas, the number of potential breeding sites is reduced compared to areas that have a lower density of buildings [21] | |
Low-density planned residential areas | WUDAPT LCZ | ||
Deprived residential areas | n/a (WUDAPT LCZ) | An increasing proportion of urban residents live in unplanned settlements where the lack of services (e.g., sanitation, drainage) and unfavourable siting (e.g., in lowlands) may contribute to the emergence of breeding sites. [68, 69]. Breeding sites are commonly found in deprived areas, including slums [8, 9, 70] | Dakar's suburbs include non-structured urbanized areas located in depressions and lowlands, with a shallow water table. These areas are prone to flooding and hence to the proliferation of larval habitats [71]. Garbage and sewage disposal services are available for most residents, but in some areas solid waste can cause clogged drains [28, 68] |
Non-residential built-up areas | WUDAPT LCZ | Breeding sites are characteristically located close to human habitations [64] and can therefore be expected to be less frequent in non-residential areas | Breeding sites tend to be located close to human dwellings rather than in uninhabited areas [28] |
Agricultural areas | Esri 2020 Land Cover, WorldCover | In general, factors that are favourable to agriculture are also favourable to the presence of larvae, e.g., lowlands, clayey or loamy soils with low runoff and the presence of water bodies [72]. Moreover, irrigation of urban and peri-urban agricultural land has led to the emergence of larval habitats [61, 73, 74] and features such as trenches, irrigation wells, water tanks, ditches etc. provide good conditions [9, 61, 67, 75]. An. gambiae is likely to develop resistance to insecticides used in agriculture [63, 76] | |
Non-agricultural vegetated areas | Esri 2020 Land Cover, WorldCover | An. gambiae usually breeds in sites where there is no upright vegetation [33] | Breeding sites tend to be located close to human dwellings rather than in uninhabited areas [28] |
Non-agricultural areas with sparse or no vegetation | Esri 2020 Land Cover, WorldCover | Breeding sites tend to be located close to human dwellings rather than in uninhabited areas [28] | |
Wetlands | Esri 2020 Land Cover, WorldCover | Wetlands/swamps are natural breeding sites often cited in the literature as potential habitats [9] | Urban wetlands are typical larval habitats [77] |
Landforms from VHR imagery | Alternative existing open product(s) | Larval habitat suitability—Sub-Saharan African cities | Larval habitat suitability—Dakar |
---|---|---|---|
Flats | Global SRTM Landforms | Topographic predictors of aquatic habitat suitability influence the spatial variation of vector density. A higher density is associated with concave planform curvature, particularly when located at the foot of steep slopes [78] | In the department of Dakar (i.e., the city centre), terrain is relatively flat, while in the suburbs, there is a succession of interdunal depressions that are prone to flooding [79] |
Peaks | |||
Ridges | |||
Shoulders | |||
Spurs | |||
Slopes | |||
Pits | |||
Valleys | |||
Foot slopes | |||
Hollows |
Continuous variables from VHR imagery | Alternative existing open product(s) | Larval habitat suitability—Sub-Saharan African cities | Larval habitat suitability—Dakar |
---|---|---|---|
Topographic Wetness Index (TWI) | Global SRTM mTPI (Multi-Scale Topographic Position Index) | TWI is a steady-state predictor of local wetness that is positively associated with the density of breeding sites [80] | Lowlands are prone to flooding during the rainy season, due to the shallow water table. Puddles can subsist in the dry season in such areas [28] |
Distance to human habitations (proxy: distance to buildings) | Calculate distance to buildings from Open Buildings | The probability of larvae presence increases when water bodies are located within 10 m of human habitations [28] | |
Distance to trees | Calculate distance to trees extracted from Esri 2020 Land Cover or WorldCover | Trees can provide refuge for digesting/resting mosquitoes [81] | |
Water pollution (proxy: distance to dumpsites) | Calculate distance to dumpsites extracted from OSM | ||
n/a | iSDAsoil: Soil pH (proxy for water pH) | A high pH (e.g., due to pollution) is not favourable for breeding and survival, natural pH is preferred [12] | An increase in pH from 7.4 to 8.2 is associated with an increase in larvae abundance [85] |
Scaling factors
Weighting factors
Aggregating criteria
Aggregating HSI to grids and validating the gridded maps
Classifying HSI into suitability classes
Hazard–b/Adult vector habitat suitability
Identifying a set of criteria (factors and constraints), and obtaining or producing the corresponding geospatial layers
Continuous variables from VHR imagery | Alternative existing open product(s) | Larval habitat suitability—Sub-Saharan African cities | Larval habitat suitability—Dakar |
---|---|---|---|
Distance to breeding sites (derived from larval habitat suitability) | n/a | In Africa, the dispersal range of Anopheles vectors of malaria from their breeding sites is generally less than 1 km and rarely exceeds 2–3 km. In peri-urban/urban areas, this range is shorter and will likely not exceed a few hundred meters when human hosts are available nearby for blood meals [8, 33, 64] | |
Distance to human dwellings (proxy: distance to buildings) | Calculate distance to Open Buildings | An. arabiensis primarily feeds and rests indoors, but due to widespread use of Long-Lasting Insecticidal Nets (LLINs) and Indoor Residual Spraying (IRS), the behaviour of this vector becomes more flexible, and it also tends to feed and rest outdoors [96] | The proximity of breeding sites to human dwellings greatly limits the spatial dispersion of vectors [28] |
Land-cover classes from VHR imagery | Alternative existing open product(s) | Larval habitat suitability—Sub-Saharan African cities | Larval habitat suitability—Dakar |
---|---|---|---|
Poorly built dwellings (proxy: low buildings) | Open buildings (no distinction based on building height) WSF3D | ||
Improved buildings (proxy: medium- and high-rise buildings) | Improved housing and mosquito proofing contribute to the decline in malaria incidence [5] | ||
Swimming pools | n/a | Swimming pools are not cited among suitable habitats | Swimming pools are not cited among suitable habitats |
Paved surface | OSM | Paved surface is not cited among suitable habitats | Paved surface is not cited among suitable habitats. The absence of vegetation implies lower vector densities [10] |
Dumpsites | OSM | Dumpsites induce the proliferation of flies, mosquitoes and rodents, and city dwellers living nearby are affected by related diseases, including malaria [58] | Vector proliferation was not observed over solid waste accumulation [85] |
Bare soil | Esri 2020 Land Cover, WorldCover | Bare soil is not cited among suitable habitats | Bare soil is not cited among suitable habitats. The absence of vegetation implies lower vector densities [10] |
Grass | Esri 2020 Land Cover, WorldCover | While mostly endophilic, An. gambiae may find suitable shady resting places in vegetation [99] | |
Trees and shrub/scrub | Esri 2020 Land Cover, WorldCover | While mostly endophilic, An. gambiae may find suitable shady resting places and nectar in trees and/or shrub. Mixing sugar meals and blood meals increases adult longevity [101]. The presence of foliage also improves the adult survival rate [33, 102]. Only female Anopheles with a high longevity can transmit P. falciparum, since the complete sporogonic cycle usually last 10–12 days in typical African climate conditions (depending on temperature and humidity) [33] | |
Water bodies | OSM, Esri 2020 Land Cover, WorldCover | Water bodies are not cited among suitable habitats | Water bodies are not cited among suitable habitats |
Shadow | n/a | Blood-fed females tend to hide themselves in humid, shady spots, either indoor or outdoor [104] |
Land-use classes from VHR imagery | Alternative existing open product(s) | Larval habitat suitability—Sub-Saharan African cities | Larval habitat suitability—Dakar |
---|---|---|---|
High-density planned residential areas | WUDAPT LCZ | Densely built-up areas in planned neighbourhoods decrease the life span of adult mosquitoes as they do not provide appropriate resting sites [102] | Areas where the presence of vegetation is important have higher vector densities than highly urbanized zones [10] |
Low-density planned residential areas | WUDAPT LCZ | More resting sites are likely to be found in low-density residential areas [102] | |
Deprived residential areas | n/a (WUDAPT LCZ) | Entomological Inoculation Rate (EIR) is higher in deprived areas located in lowlands with a shallow water table [71] | |
Non-residential areas | WUDAPT LCZ | An. gambiae being highly anthropophilic and biting between sunset and sunrise [33], non-residential areas do not generally provide suitable conditions for adult survival | An. arabiensis is highly anthropophilic [71] and uninhabited areas are thus unlikely to provide suitable conditions for adult survival |
Agricultural areas | Esri 2020 Land Cover, WorldCover | Urban agricultural areas are one of the key elements in malaria risk mapping in cities. Market gardens increase the availability of resting sites [102] | Market gardens provide suitable resting sites to adults [34] |
Non-agricultural vegetated areas | Esri 2020 Land Cover, WorldCover | An. gambiae being highly anthropophilic and biting between sunset and sunrise [33], uninhabited areas do not provide suitable conditions for adult survival | An. arabiensis is highly anthropophilic [71] and uninhabited areas are thus unlikely to provide suitable conditions for adult survival |
Non-agricultural areas with sparse or no vegetation | Esri 2020 Land Cover, WorldCover | ||
Wetlands | Esri 2020 Land Cover, WorldCover |
Scaling factors
Weighting factors, aggregating criteria, aggregating HSI to grid, verifying, classifying into suitability classes
Population and vulnerability
Urban malaria exposure
Results
Hazard–a/larval habitat suitability
LC classes | Score | LU classes | Score | Landforms | Score |
---|---|---|---|---|---|
Buildings | 0 | High-density planned residential areas | 9 | Flats | 17 |
Swimming pools | 8 | Low-density planned residential areas | 16 | Peaks | 0 |
Paved surface | 5 | Deprived residential areas | 79 | Ridges | 1 |
Dumpsites | 10 | Non-residential built-up areas | 0 | Shoulders | 7 |
Bare soil | 16 | Agricultural areas | 87 | Spurs | 7 |
Grass | 19 | Non-agricultural vegetated areas | 36 | Slopes | 2 |
Shrubs | 11 | Non-agricultural areas with sparse or no veg | 31 | Pits | 100 |
Trees | 5 | Wetlands | 100 | Valleys | 70 |
Small water bodies | 100 | Footslopes | 53 | ||
Medium water bodies | 88 | Hollows | 42 | ||
Large water bodies | 54 | ||||
Water courses | 11 | ||||
Marine waters | 9 | ||||
Shadow | 8 |
LC classes | Score | LU classes | Score |
---|---|---|---|
Low buildings (incl. poorly built) | 100 | High-density planned residential areas | 35 |
Medium and high-rise buildings | 39 | Low-density planned residential areas | 20 |
Swimming pools | 8 | Deprived residential areas | 100 |
Paved surface | 0 | Non-residential built-up areas | 0 |
Dump sites | 26 | Agricultural areas | 29 |
Bare soil | 2 | Non-agricultural vegetated areas | 17 |
Grass | 35 | Non-agricultural areas with sparse or no veg | 7 |
Trees and shrubs | 52 | Wetlands | 30 |
Water bodies | 18 | ||
Shadow | 18 |