Drone-based composite risk mapping reveals vegetation’shade interaction and housing typology as key determinants of Aedes habitat risk - Scientific Reports

Persistent dengue transmission in tropical cities reflects a complex interplay between environmental microclimates and urban housing structure that supports Aedes mosquito breeding. A study published in Scientific Reports applies drone-based microhabitat risk mapping, integrated with a biologically defined Composite Risk Index, to quantify the fine-scale environmental drivers of that risk across residential typologies in Section 24, Shah Alam, Malaysia.

Two indices from ordinary RGB imagery

High-resolution RGB imagery captured with a DJI Phantom 4 Pro was processed to derive the Brightness Index as a proxy for shade intensity, and the Excess Green Index as an indicator of vegetation density. Both were integrated a priori into a Composite Risk Index designed to operationalise the ecological conditions already known to favour Aedes.

Spatial analysis revealed a consistent risk gradient, with terrace housing showing higher CRI values than flat complexes: low-density terrace, then dense terrace, then medium-rise, then high-rise. In other words, housing typology modulates how microhabitat risk is expressed spatially — vegetation presence alone does not explain it.

Model calibration showed high predictive agreement, with an R² of 0.91, and the top 20 per cent of CRI-ranked pixels captured 65 per cent of observed breeding-prone zones.

Why existing surveillance misses these sites

Dengue remains one of the fastest-expanding mosquito-borne diseases in the world, with over 400 million infections estimated annually across more than 120 countries. In Malaysia it has shifted from cyclical outbreaks to a persistent public health threat, driven by rapid urbanisation, environmental degradation and the spread of Aedes aegypti and Aedes albopictus.

Despite decades of vector control, the disease continues to intensify in urban and suburban areas where housing, built environment and vegetation coexist in complex spatial arrangements. Traditional entomological surveillance relying on larval indices and manual inspection is constrained by labour, cost and temporal coverage, leaving gaps in identifying high-risk micro-environments.

What drone imagery adds over satellites

High-resolution UAV imagery allows fine-scale mapping of environmental features that influence vector ecology, including vegetation structure, shade intensity and surface moisture proxies — the parameters that shape the microclimatic conditions determining Aedes survival, reproduction and dispersal.

Conventional satellite imagery lacks the spatial resolution to capture neighbourhood-scale heterogeneity. Drone acquisition provides centimetre-level precision, which is what makes potential breeding habitats embedded inside residential areas detectable at all.

In tropical cities such as Shah Alam, Aedes breeding persists even in well-maintained neighbourhoods, suggesting shade-vegetation interactions matter more than previously assumed. Vegetation retains the humidity and organic matter larvae need, while shade stabilises temperature and prolongs water availability in artificial containers. Few studies have quantitatively examined how those factors interact spatially across contrasting housing types; most have relied on coarse land-use classifications or limited ground surveys that miss the gradients sustaining mosquito populations.

The study area

The work was conducted in Shah Alam, Selangor, about 25 km southwest of Kuala Lumpur in the Klang Valley conurbation. Selangor is Malaysia's most urbanised and dengue-endemic state, with mean daily temperatures between 26°C and 32°C and annual rainfall exceeding 2,400 mm — conditions supporting year-round breeding. Section 24 was chosen for its heterogeneous urban morphology and recurring dengue notifications recorded by the Shah Alam City Council.

The area comprises four residential typologies: a high-rise complex, a medium-rise block, dense terrace housing and low-density terrace housing. The flat complexes are dominated by multi-storey buildings with extensive impervious surfaces, minimal green space and narrow perimeter corridors that generate transient shaded micro-sites but little organic substrate.

The terrace zones, by contrast, contain single- and double-storey houses surrounded by gardens, boundary hedges and rear-yard vegetation clusters that trap humidity and reduce solar exposure. The low-density terrace zone has extensive vegetated patches within and around residential parcels, providing the contrasting landscape context for vegetation quantification.

By combining drone photogrammetry, GIS-based spatial modelling and ground validation, the authors present the framework as a reproducible basis for precision vector control aligned with Malaysia's Integrated Vector Management strategy.