Cemetery and burial ground expansion monitoring
High-resolution satellite imagery can track cemetery expansion, encroachment disputes and new grave-plot formation at sub-metre scale, giving land administrators an auditable spatial record where ground surveys are impractical.
Sensors
- Airbus Pléiades Neo: 30 cm panchromatic, 50 cm multispectral (blue, green, red, red-edge, NIR); revisit up to twice daily at mid-latitudes with the four-satellite constellation. At this resolution individual grave plots, kerb stones and access paths are directly measurable. The red-edge band helps separate recently disturbed soil from established turf cover.
- Maxar WorldView Legion: 30 cm panchromatic resolution; six satellites provide revisit of up to 15 times per day over priority targets. Useful for rapid detection of boundary-wall construction or encroachment events that unfold over days rather than months.
- Planet SkySat: 50 cm panchromatic, approximately 1 m multispectral; tasked revisit. Less sharp than Pléiades Neo or Legion but cost-effective for monitoring larger cemetery complexes or tracking slow, incremental expansion over quarterly intervals.
- Sentinel-2 MSI: 10 m multispectral (visible, NIR, SWIR) at 5-day global revisit, freely available. Too coarse to resolve individual plots, but NDVI and bare-soil indices computed from Sentinel-2 time series reliably flag the spectral signature of freshly disturbed ground at the scale of a new cemetery section, providing a low-cost screening layer before commercial tasking.
Why cemeteries are a land-administration problem satellites can actually solve
In many cities, cemeteries are among the fastest-growing land consumers that planners systematically under-monitor. Manila's Rizal Memorial complex, Jakarta's Tanah Kusir and Lagos's Ikoyi Cemetery are documented examples of burial grounds that have expanded incrementally over decades, often without corresponding updates to cadastral records. The pattern is consistent: a formal boundary exists on paper, physical occupation extends beyond it, and the discrepancy compounds until a dispute or a development project forces a reckoning.
Satellite change detection offers something a periodic ground survey cannot: a continuous, dated spatial record. With archive imagery from Pléiades Neo or WorldView Legion going back several years, an analyst can reconstruct the precise sequence of encroachment, identify when a boundary wall was extended and measure how many square metres of adjacent land were absorbed. That record is admissible as evidence in land tribunals in a growing number of jurisdictions.
What a floating roof gives away: the spectral signature of a new grave
A freshly dug grave is spectrally similar to any other disturbed-soil feature: a construction trench, a drainage ditch, a road cut. In the visible and near-infrared, bare mineral soil reflects more strongly than vegetated ground, and that contrast is detectable in Sentinel-2 NDVI composites at the scale of a newly opened cemetery section. The problem is discrimination. At 10 m resolution, a single grave plot is sub-pixel. You need commercial imagery to resolve individual features.
At 30 to 50 cm resolution, the geometry becomes diagnostic. Grave plots in most cultural traditions are laid out in regular rows with consistent orientation and spacing. That regularity is detectable by template-matching or object-based image analysis even when individual mounds are only 20 to 40 cm high. The mounds themselves are low-relief earthworks, and this is where solar geometry matters. Early-morning or late-afternoon acquisitions, when the sun angle is below 30 degrees, cast shadows long enough to make 20 cm relief visible in the panchromatic band. A midday acquisition of the same scene may show almost nothing. Tasking requests for cemetery monitoring should specify a sun elevation below 35 degrees wherever the satellite's agility permits.
Grave markers, kerb stones and concrete grave surrounds add a second detection cue: their spectral contrast with surrounding soil or turf is high in the visible bands, and their geometric regularity distinguishes them from random debris. Pléiades Neo's red-edge band (at roughly 700 nm) is particularly useful here because it separates sparse, stressed grass on older plots from bare soil on new ones, helping to date expansion phases without relying solely on archive comparisons.
Access roads and boundary walls: the infrastructure that precedes the plots
Cemetery expansion rarely begins with graves. It begins with infrastructure. A new access path is cut, a section of boundary wall is extended or a drainage channel is dug. These features are larger and spectrally more distinct than individual grave mounds, and they appear weeks or months before the first burials in a new section. Monitoring for these precursor features gives land administrators lead time.
At 30 cm resolution, a compacted-earth access path 1.5 m wide is a clearly resolved linear feature. Concrete or masonry boundary walls cast measurable shadows even at moderate sun angles, and their linear geometry is straightforward to extract with edge-detection algorithms. A change-detection workflow that flags new linear features at a cemetery boundary is therefore more sensitive, and more actionable, than one that waits for grave mounds to accumulate.
Honest limits: what the imagery cannot tell you
Sub-metre imagery resolves grave plots; it does not resolve occupancy. A plot that appears freshly dug may be a family-reserved grave prepared years before use. Conversely, a section that looks undisturbed from above may have had subsurface interments if the surface was carefully reinstated. Remote sensing maps surface evidence, not subsurface fact.
Cloud cover is a persistent constraint in tropical cities, where many of the world's most congested cemeteries are located. A single Pléiades Neo or SkySat acquisition may be cloud-obscured, and unlike Sentinel-2, commercial tasking does not automatically retry. A monitoring programme needs to budget for multiple tasking attempts per epoch, particularly during monsoon seasons. Expect usable clear-sky imagery on perhaps 40 to 60 per cent of tasked dates in persistently cloudy regions. This is not a reason to abandon satellite monitoring; it is a reason to plan acquisition schedules realistically.
Shadow analysis for low-relief earthworks requires deliberate tasking geometry. If a client submits a standard tasking request without specifying sun-angle constraints, the satellite operator will often acquire at whatever geometry is convenient. The resulting midday image may show a flat, featureless surface where a carefully timed acquisition would have revealed rows of mounds. This is a workflow detail, not a sensor limitation, but it is one that matters enormously for this specific use case.
From pixels to a land-administration record
The analytic output for a cemetery monitoring programme is typically a set of georeferenced polygon layers: one for the formal cadastral boundary, one for the observed physical extent at each monitoring epoch, and one for features identified as new grave plots, access infrastructure or encroachment structures. Overlaying these against a municipal cadastral dataset immediately quantifies the discrepancy between registered and occupied land.
Change magnitude can be expressed as area (square metres of new burial ground per year) or as plot count (estimated number of new graves per epoch, derived from object-based counting at sub-metre resolution). Plot-count estimation carries uncertainty, typically plus or minus 10 to 20 per cent depending on plot regularity and image quality, and should be reported as a range rather than a point figure.
Satellize has built similar disturbed-soil and land-cover change workflows for the Kingdom of Tonga crop-estimation programme, where spectral separation of bare soil, crop cover and fallow ground is a closely related analytical problem. The same Sentinel-2 screening layer combined with commercial tasking for confirmation translates directly to cemetery monitoring at municipal scale. A city land office wanting to commission a baseline survey and quarterly monitoring programme should begin by defining the boundary of interest and specifying the archive depth needed to support any existing dispute.
Typical figures
| Best spatial resolution (panchromatic) | 30 cm (Pléiades Neo, WorldView Legion) |
| Best spatial resolution (multispectral) | 50 cm (Pléiades Neo); ~1 m (SkySat) |
| Screening layer resolution | 10 m (Sentinel-2 MSI, freely available) |
| Revisit for tasked commercial imagery | Up to twice daily (Pléiades Neo); up to 15× daily (WorldView Legion over priority targets) |
| Minimum detectable earthwork relief (shadow method) | ~20 cm at sun elevation <35°; effectively undetectable at sun elevation >55° |
| Minimum resolvable linear feature (access path) | ~1 m width at 30 cm resolution |
| Archive depth (commercial sensors) | Pléiades 1A/1B from 2012; WorldView-1/2/3 from 2007–2009; SkySat from ~2014 |
| Spectral bands used | Panchromatic; blue, green, red, red-edge (~700 nm), NIR; SWIR (Sentinel-2 for soil index) |
| Typical delivery format | Georeferenced GeoTIFF imagery; GeoPackage or Shapefile polygon layers; PDF change report |
| Cloud-affected acquisition risk (tropical regions) | 40–60% of tasked dates may be cloud-obscured; multi-attempt scheduling required |
Analytics Satellize can run
| Cemetery boundary discrepancy map | Manual digitisation and object-based image analysis (OBIA) of physical extent versus registered cadastral boundary | Polygon GIS layer showing formal boundary, observed physical extent and encroachment area in m² |
| New grave-plot count per epoch | Template-matching and OBIA on sub-metre panchromatic imagery; row-regularity filtering to exclude non-grave disturbed soil | Estimated plot count with ±10–20% uncertainty range; tabular report per monitoring period |
| Disturbed-soil screening alert | Sentinel-2 bare-soil index (BSI) and NDVI time-series anomaly detection at 10 m; flags candidate expansion zones for commercial tasking | Automated alert with bounding box of flagged area; triggers tasking request |
| Shadow-derived low-relief earthwork map | Shadow length measurement in low-sun-angle panchromatic imagery; height estimation from known solar geometry and image metadata | Raster layer of estimated surface relief; polygon overlay of probable grave-mound rows |
| Infrastructure precursor detection | Linear feature extraction (edge detection, morphological filtering) applied to change-differenced imagery at cemetery boundary | GIS layer of new linear features (paths, walls, drainage) with date of first appearance |
| Multi-epoch expansion timeline | Bi-temporal and multi-date change detection across archive imagery stack; area measurement per epoch | Time-series chart of cemetery footprint growth in m² per year; annotated image mosaic for land tribunal use |
Who does the work
We can get this done for you. Satellize runs its own analyst desk and a strong science team. You do not buy a data feed and work out what it means; our people source the imagery, run the analysis described on this page, and hand you the answer with its confidence limits stated. Discuss this requirement.