Skip to content
Back to search results

Dasymetric Mapping

Introduced by
J. K. Wright (introduced 1936); modern surface method by Jeremy Mennis
Year
2003
Type
Cartographic areal-interpolation technique using ancillary data
Family
Areal interpolation / thematic mapping
Font size

What does it do?

Dasymetric mapping is a cartographic and areal-interpolation technique that redistributes data reported for arbitrary administrative zones — such as census counts — onto more meaningful boundaries derived from ancillary information about where the phenomenon actually occurs.

Instead of pretending population is spread evenly across a census tract, it uses land cover or land use to push people into the residential parts and out of lakes, parks, and industry, producing a far more realistic population surface while preserving each zone's reported total.

The intuition behind the method

A choropleth map shades a whole county one colour, implying everyone is spread uniformly across it — including across the reservoir and the airport. That is obviously wrong. Dasymetric mapping fixes it by overlaying a layer that tells you where people can live: it then crowds the county's known population into the residential land and empties the uninhabitable land, so the map shows high density where there are houses and zero where there is water or forest. Crucially, the people moved out of one part are added to another within the same zone, so the county still totals exactly what the census reported.

History

The term and idea of the dasymetric map were introduced by the American cartographer John K. Wright in 1936, who redistributed Cape Cod's population using knowledge of where it actually lived rather than uniform township shading. The method languished for decades until GIS and digital land-cover data made it routine. Tobler's 1979 pycnophylactic interpolation supplied the volume-preserving principle, and Jeremy Mennis's 2003 paper in The Professional Geographer systematized the modern approach of using classified land cover to generate population surface models, which is now standard practice in population mapping.

When to use it

Use dasymetric mapping when you have data aggregated to coarse, arbitrary zones but need a more realistic spatial distribution — to map population density honestly, to estimate counts for areas that do not align with the reporting zones (areal interpolation), or to combine demographic data with environmental or hazard layers. It requires ancillary data that genuinely correlates with the phenomenon's location, such as land cover for population. It is unsuitable when no relevant ancillary layer exists, when the ancillary data are coarser or less current than the source zones, or when the assumed within-class homogeneity of density is badly violated.

Limitations and considerations

  • Wholly dependent on the quality, resolution, currency, and relevance of the ancillary data.
  • Assumes density is homogeneous within each ancillary class, which is only an approximation.
  • Setting class density weights can be subjective unless they are empirically estimated.
  • Errors in the ancillary layer propagate directly into the redistributed surface, sometimes invisibly.

Source conditions do not replace independent scientific review or validation by an analysis engine.

Strengths

  • Produces far more realistic population (or other) surfaces than uniform choropleth or simple areal weighting.
  • Volume-preserving: zonal totals are conserved, keeping results consistent with official statistics.
  • Enables interpolation between incompatible zone systems and integration with environmental or risk layers.
  • Flexible in sophistication, from quick binary (habitable/uninhabitable) masks to empirically calibrated multi-class models.

Applications

  • Producing realistic gridded population density surfaces from census tract or county data.
  • Estimating the population at risk within flood zones, hazard buffers, or service catchments that cross census boundaries.
  • Interpolating data between mismatched zone systems, such as old and new census geographies.
  • Integrating demographic counts with remotely sensed land cover for exposure and vulnerability analysis.
  • Building inputs to global gridded population datasets and to environmental epidemiology studies.

How it works

How it works

1

The source zones (e.g. census tracts) are overlaid with an ancillary layer — typically land cover or land use classified into categories such as high-density residential, low-density residential, commercial, and uninhabitable. The overlay partitions each zone z into sub-areas A_zc, the area of each ancillary class c falling within that zone.

Formula
\text{zone } z = \bigcup_c A_{zc},\quad A_{zc} = \text{area of class } c \text{ in zone } z
2

Each ancillary class is assigned a relative population density weight d_c reflecting how intensely people occupy that land type — high for dense residential, low for sparse, zero for water or forest. Weights can be set from expert judgment (binary or three-class methods), from a sample of zones, or estimated empirically by regressing zonal counts on class areas.

Formula
\hat{d}_c = \text{relative population density assigned to class } c
3

The zone's reported total P_z is allocated to its ancillary sub-areas in proportion to each sub-area's weighted area, d_c times A_zc, divided by the zone's total weighted area. Dense residential land thus receives a large share of the population and uninhabitable land receives little or none, yielding a population count P_zc for every sub-area.

Formula
P_{zc} = P_z\, \frac{\hat{d}_c\, A_{zc}}{\sum_{c'} \hat{d}_{c'}\, A_{zc'}}
4

The redistribution is volume-preserving (pycnophylactic): the sub-area counts within each source zone sum exactly back to the zone's original total. This guarantees that dasymetric mapping rearranges where the count sits without inventing or destroying any of it, so the result remains consistent with the published statistics.

Formula
\sum_c P_{zc} = P_z\quad \forall z

Common pitfalls

  • Using an ancillary layer that is older or coarser than the source data, so it misrepresents current land use.
  • Treating broad land-cover classes as if density were uniform within them when it varies widely.
  • Forgetting to enforce the pycnophylactic constraint, producing counts that no longer match the source totals.
  • Over-trusting the smooth, detailed-looking output as ground truth when it is still a model built on assumptions.

Frequently asked questions

How is dasymetric mapping different from a choropleth map?

A choropleth shades each reporting zone a single tone, implicitly assuming the value is uniform across the whole zone, including uninhabitable land. Dasymetric mapping breaks each zone into sub-areas using ancillary data and redistributes the value so that it concentrates where the phenomenon actually occurs, producing internally varying density while conserving the zone's total. It is a more truthful depiction of an unevenly distributed quantity.

What does 'pycnophylactic' (volume-preserving) mean here?

It means the redistribution conserves mass: the counts allocated to all the sub-areas inside a source zone add back up to exactly that zone's reported total, with nothing created or lost. This constraint keeps the dasymetric surface consistent with the published statistics, distinguishing it from interpolation methods that merely smooth values without honouring the original totals.

What ancillary data are typically used for population dasymetric mapping?

Classified land cover or land use is the most common ancillary source — for example the National Land Cover Database — distinguishing residential land of varying density from commercial, industrial, water, forest, and other uninhabitable categories. Imperviousness, building footprints, road networks, and night-time lights are also used. The key requirement is that the ancillary layer reliably indicates where the population resides.

Related methods

Sources

  1. Mennis, J. (2003). Generating surface models of population using dasymetric mapping. The Professional Geographer, 55(1), 31–42.

    Open DOI reference
    Copied.
  2. Kraak, M.-J., & Ormeling, F. J. (2010). Cartography: Visualization of Geospatial Data (3rd ed.). Prentice Hall.

    ISBN: 9780273722793
    Copied.