dasymetric

The goal of dasymetric mapping is to display statistical data (like census data) in meaningful spatial zones.

Installation

And the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("JaFro96/dasymetric")

Example

As a case study we try to predict population counts for each district of Münster (Westfalen) using land cover data as ancillary information.

# plot population of 2018
require(sf)
load("data/population_counts.rda")
plot(population_counts["population"],breaks = c(0,5000,10000,15000,20000,25000,30000,35000,40000,45000), main="Population (2018)")

Below the dasymetric map is plotted which exhibits similar patterns in the population distribution

require(dasymetric)
load("data/corine_18.rda")
urban = prep_landuse(corine_18)

# source geometry covering entire Münster 
source_geom = st_union(population_counts)
# add population of Münster
source = st_sf(ID = 1, pop_sum = sum(population_counts["population"]$population), source_geom)

# dasymetric map with landuse information as ancillary data
dm_pop = dasymetric_map(population_counts, source, urban, extensive = "pop_sum")
plot(dm_pop["pop_sum"],breaks = c(0,5000,10000,15000,20000,25000,30000,35000,40000,45000),main="Dasymetric Population Map based on Land Use Information (2018)")

… contrary to the population distribution using area-weighted interpolation:

require(areal)
# Area-weighted interpolation of Münsters districts
aw_pop = aw_interpolate(population_counts,NR_STATIST,source = source, sid = ID,weight = "sum", extensive = "pop_sum", output = "sf")
plot(aw_pop["pop_sum"],breaks = c(0,5000,10000,15000,20000,25000,30000,35000,40000,45000),main="Area-weighted Interpolation of Population (2018)")

Data Sources

• CORINE Land Cover 5 ha –> © GeoBasis-DE / BKG (2021)

• district boundaries –> opendata.stadt-muenster.de

• population counts –> opendata.stadt-muenster.de