The BioDT Recreational Potential Model for Scotland

Worked example

Authors

Affiliation

Joe Marsh Rossney

UK Centre for Ecology & Hydrology

Maddalena Tigli

Chris Andrews

Jan Dick

Simon Rolph

Abstract

This notebook demonstrates basic usage of the Recreational Potential model. It is assumed that the user is familiar with installing R packages, executing lines of R code in interactive sessions, and running R scripts.

Getting started

The Recreational Potential model has been developed on recent (2025) versions of R — specifically versions 4.4.2 up to 4.5.0. We cannot guarantee that it will work for versions of R outside of this range.

The following command tells you what version is currently active:

# Ideally this should be between 4.4.2 and 4.5.0
R.version.string

[1] "R version 4.5.0 (2025-04-11)"

Certain C++ libraries are required for the model to work, most importantly related to GDAL. See the terra documentation for guidance.

It is recommended to perform the following steps using an R environment managed by renv.

Ensure you have either remotes or devtools installed (using e.g. install.packages or renv::install).

install.packages("remotes")

The following package(s) will be installed:
- remotes [2.5.0]
These packages will be installed into "~/github.com/BioDT/ces-recreation-reports/renv/library/linux-ubuntu-jammy/R-4.5/x86_64-pc-linux-gnu".

# Installing packages --------------------------------------------------------
- Installing remotes ...                        OK [linked from cache]
Successfully installed 1 package in 4.4 milliseconds.

Now install the package itself. We will target version 1.0.¹

remotes::install_github("BioDT/uc-ces-recreation", ref="v1.0")

Using GitHub PAT from the git credential store.

Skipping install of 'biodt.recreation' from a github remote, the SHA1 (c5f60e30) has not changed since last install.
  Use `force = TRUE` to force installation

Next, you will need to download the input data. The total size is over 2GB, so it is important to set the timeout option to something generous.

options(timeout=1200)
biodt.recreation::download_data()

Skipping download of SLSRA since /home/joe/github.com/BioDT/ces-recreation-reports/renv/library/linux-ubuntu-jammy/R-4.5/x86_64-pc-linux-gnu/biodt.recreation/extdata/rasters/Scotland/SLSRA.tif already exists.

Skipping download of FIPS_N since /home/joe/github.com/BioDT/ces-recreation-reports/renv/library/linux-ubuntu-jammy/R-4.5/x86_64-pc-linux-gnu/biodt.recreation/extdata/rasters/Scotland/FIPS_N.tif already exists.

Skipping download of FIPS_I since /home/joe/github.com/BioDT/ces-recreation-reports/renv/library/linux-ubuntu-jammy/R-4.5/x86_64-pc-linux-gnu/biodt.recreation/extdata/rasters/Scotland/FIPS_I.tif already exists.

Skipping download of Water since /home/joe/github.com/BioDT/ces-recreation-reports/renv/library/linux-ubuntu-jammy/R-4.5/x86_64-pc-linux-gnu/biodt.recreation/extdata/rasters/Scotland/Water.tif already exists.

Where are the files?

Do not be alarmed if you cannot see any new files appearing in your working directory. When download_data is called without arguments, the data will be downloaded to the default location which is inside the directory in which the package is installed (under the subdirectory extdata/rasters/Scotland/). You may not know where on your filesystem packages are installed, but that’s ok — the package knows where to look for these files.

The preset personas

We first demonstrate the RP Model applied to an area surrounding the Bush estate near Edinburgh² using two contrasting personas:

1. Hard Recreationalist: this persona values remote, challenging environments and avoids built-up or highly managed areas.
2. Soft Recreationalist: this persona values tranquil, accessible landscapes and natural water features, and avoids steep or harsh terrain.

The resulting maps will illustrate how different preferences lead to different areas being highlighted for recreational value.

These personas are provided in the model as ‘presets’; users may customise them or simply use them out-of-the-box.

First, let’s save ourselves some typing of biodt.recreation::<function> by exposing all of the functions in the biodt.recreation namespace.

library(biodt.recreation)

Now we will load the Hard Recreationalist and Soft Recreationalist personas from the presets file.

persona_file <- get_preset_persona_file()

hard_recreationalist <- load_persona(persona_file, "Hard_Recreationalist")

Loading persona 'Hard_Recreationalist' from file '/home/joe/github.com/BioDT/ces-recreation-reports/renv/library/linux-ubuntu-jammy/R-4.5/x86_64-pc-linux-gnu/biodt.recreation/extdata/personas/presets.csv'

soft_recreationalist <- load_persona(persona_file, "Soft_Recreationalist")

Loading persona 'Soft_Recreationalist' from file '/home/joe/github.com/BioDT/ces-recreation-reports/renv/library/linux-ubuntu-jammy/R-4.5/x86_64-pc-linux-gnu/biodt.recreation/extdata/personas/presets.csv'

The tab panel below is helpful for inspecting and comparing the two preset personas.

persona_table <- function(persona) {

  # Construct named vector of persona descriptions
  config <- load_config()
  descriptions <- stats::setNames(config$Description, config$Name)

  # Align order (just in case)
  descriptions <- descriptions[names(persona)]

  # Construct data.frame from the two named vectors
  df <- data.frame(
    Item = names(persona),
    Description = as.vector(descriptions),
    Score = as.vector(persona)
  )

  # Convert dataframe to HTML table
  knitr::kable(df)
}

persona_hist <- function(persona) {

  # Make 2x2 grid for histograms
  par(mfrow = c(2, 2))

  for (component in c("SLSRA", "FIPS_N", "FIPS_I", "Water")) {

    # Scores for this component
    persona_subset <- persona[grep(paste0("^", component), names(persona))]

    # Construct histogram with 11 equally space bins centred on
    # the integers 0, 1, ..., 10.
    hist(
      persona_subset,
      main=component,
      breaks=seq(-0.5, 10.5, by=1),
      xlab="Score",
      ylab="Frequency"
    )
  }
}

Hard Recreationalist

persona_hist(hard_recreationalist)

persona_table(hard_recreationalist)

Item	Description	Score
FIPS_I_LocalPathNetwork_2	Path	5
FIPS_I_NationalCycleNetwork_1	On Road: Paved Surface	2
FIPS_I_NationalCycleNetwork_2	Traffic Free: Unpaved Surface	8
FIPS_I_NationalCycleNetwork_3	Traffic Free: Paved Surface	6
FIPS_I_NationalCycleNetwork_4	On Road: Unpaved Surface	7
FIPS_I_RoadsTracks_1	Motorway	0
FIPS_I_RoadsTracks_2	A Road	0
FIPS_I_RoadsTracks_3	B Road	0
FIPS_I_RoadsTracks_4	Minor or local road	0
FIPS_I_RoadsTracks_5	Access roads or Track	0
FIPS_N_Landform_1	Foothills	7
FIPS_N_Landform_10	Rock Walls	10
FIPS_N_Landform_11	Uplands	7
FIPS_N_Landform_12	Valley sides	7
FIPS_N_Landform_13	Valley bottom	3
FIPS_N_Landform_14	Built-up areas	1
FIPS_N_Landform_15	Saltings	0
FIPS_N_Landform_16	Hummocks, mounds or moraines	3
FIPS_N_Landform_2	Mountains	10
FIPS_N_Landform_3	Terraces	5
FIPS_N_Landform_4	Flood plain	1
FIPS_N_Landform_5	Beaches or Dunes	7
FIPS_N_Landform_6	Rocks or Scree	6
FIPS_N_Landform_7	Depressions	1
FIPS_N_Landform_8	Hills	6
FIPS_N_Landform_9	Lowlands	3
FIPS_N_Slope_1	No slope	1
FIPS_N_Slope_2	Gentle slope	1
FIPS_N_Slope_3	Medium slope	2
FIPS_N_Slope_4	Steep slope	4
FIPS_N_Slope_5	Very steep slope	7
FIPS_N_Slope_6	Extremely steep slope	8
FIPS_N_Soil_1	Peat or Organic	6
FIPS_N_Soil_2	Mineral	2
SLSRA_CP_2	Country Park	1
SLSRA_HNV_2	Designated High Nature Value (HNV) farmland	4
SLSRA_LCM_1	Alpine and subalpine grassland	8
SLSRA_LCM_10	Dry grassland	5
SLSRA_LCM_11	Freshwater	7
SLSRA_LCM_12	Inland cliffs, rock pavements and outcrops	10
SLSRA_LCM_13	Lines of trees, small planted woodlands, early-stage woodland and coppice	2
SLSRA_LCM_14	Littoral sediment or saltmarsh	2
SLSRA_LCM_15	Mesic grassland	4
SLSRA_LCM_16	Mixed deciduous and coniferous woodland	5
SLSRA_LCM_17	Non-native coniferous plantation	2
SLSRA_LCM_18	Raised and blanket bog	1
SLSRA_LCM_19	Riverine and fen scrubs	2
SLSRA_LCM_2	Arable land and market gardens	2
SLSRA_LCM_20	Rock cliffs, ledges and shores	10
SLSRA_LCM_21	Scots pine woodland	7
SLSRA_LCM_22	Screes	5
SLSRA_LCM_23	Seasonally wet and wet grassland	5
SLSRA_LCM_24	Temperate montane scrub	7
SLSRA_LCM_25	Temperate shrub heathland	7
SLSRA_LCM_26	Valley mires, poor fens and transition mires	1
SLSRA_LCM_27	Windthrown woodland	1
SLSRA_LCM_28	Woodland fringes and clearings and tall forb stands	1
SLSRA_LCM_3	Arctic, alpine and subalpine scrub	8
SLSRA_LCM_4	Bare field or exposed soil	1
SLSRA_LCM_5	Base-rich fens and calcareous spring mires	2
SLSRA_LCM_6	Broadleaved deciduous woodland	7
SLSRA_LCM_7	Built-up area	1
SLSRA_LCM_8	Coastal dunes and sandy shore	7
SLSRA_LCM_9	Coastal shingle	3
SLSRA_NNR_2	National Nature Reserve (NNR)	6
SLSRA_NP_2	National Park	10
SLSRA_NR_2	Nature Reserve	3
SLSRA_RP_2	Regional Park	2
SLSRA_RSPB_2	Royal Society for the Protection of Birds (RSPB) Reserve	1
SLSRA_SAC_2	Special Area of Conservation (SAC)	3
SLSRA_SPA_2	Special Protection Area (SPA)	3
SLSRA_SSSI_2	Site of Special Scientific Interest (SSSI)	2
SLSRA_SWT_2	Scottish Wildlife Trust Reserve	2
SLSRA_WLA_2	Wild Land Areas	7
Water_Lakes_1	Pond	1
Water_Lakes_2	Lochan	3
Water_Lakes_3	Small Lochs	5
Water_Lakes_4	Medium Lochs	4
Water_Lakes_5	Large Lochs	7
Water_Lakes_6	Major Lochs	7
Water_Rivers_1	Minor river or tributary	3
Water_Rivers_2	Unnamed minor stream or tributary	3
Water_Rivers_3	Major river or tributary	7
Water_Rivers_4	Named minor stream or tributary	3
Water_Rivers_5	Lake	5
Water_Rivers_6	Tidal river or estuary	4
Water_Rivers_7	Canal	3

Soft Recreationalist

persona_hist(soft_recreationalist)

persona_table(soft_recreationalist)

Item	Description	Score
FIPS_I_LocalPathNetwork_2	Path	7
FIPS_I_NationalCycleNetwork_1	On Road: Paved Surface	5
FIPS_I_NationalCycleNetwork_2	Traffic Free: Unpaved Surface	6
FIPS_I_NationalCycleNetwork_3	Traffic Free: Paved Surface	10
FIPS_I_NationalCycleNetwork_4	On Road: Unpaved Surface	5
FIPS_I_RoadsTracks_1	Motorway	0
FIPS_I_RoadsTracks_2	A Road	0
FIPS_I_RoadsTracks_3	B Road	3
FIPS_I_RoadsTracks_4	Minor or local road	5
FIPS_I_RoadsTracks_5	Access roads or Track	8
FIPS_N_Landform_1	Foothills	8
FIPS_N_Landform_10	Rock Walls	1
FIPS_N_Landform_11	Uplands	7
FIPS_N_Landform_12	Valley sides	2
FIPS_N_Landform_13	Valley bottom	7
FIPS_N_Landform_14	Built-up areas	1
FIPS_N_Landform_15	Saltings	0
FIPS_N_Landform_16	Hummocks, mounds or moraines	3
FIPS_N_Landform_2	Mountains	4
FIPS_N_Landform_3	Terraces	3
FIPS_N_Landform_4	Flood plain	2
FIPS_N_Landform_5	Beaches or Dunes	9
FIPS_N_Landform_6	Rocks or Scree	1
FIPS_N_Landform_7	Depressions	1
FIPS_N_Landform_8	Hills	4
FIPS_N_Landform_9	Lowlands	8
FIPS_N_Slope_1	No slope	9
FIPS_N_Slope_2	Gentle slope	8
FIPS_N_Slope_3	Medium slope	6
FIPS_N_Slope_4	Steep slope	4
FIPS_N_Slope_5	Very steep slope	3
FIPS_N_Slope_6	Extremely steep slope	1
FIPS_N_Soil_1	Peat or Organic	3
FIPS_N_Soil_2	Mineral	5
SLSRA_CP_2	Country Park	8
SLSRA_HNV_2	Designated High Nature Value (HNV) farmland	5
SLSRA_LCM_1	Alpine and subalpine grassland	2
SLSRA_LCM_10	Dry grassland	3
SLSRA_LCM_11	Freshwater	9
SLSRA_LCM_12	Inland cliffs, rock pavements and outcrops	3
SLSRA_LCM_13	Lines of trees, small planted woodlands, early-stage woodland and coppice	6
SLSRA_LCM_14	Littoral sediment or saltmarsh	4
SLSRA_LCM_15	Mesic grassland	5
SLSRA_LCM_16	Mixed deciduous and coniferous woodland	9
SLSRA_LCM_17	Non-native coniferous plantation	7
SLSRA_LCM_18	Raised and blanket bog	5
SLSRA_LCM_19	Riverine and fen scrubs	4
SLSRA_LCM_2	Arable land and market gardens	4
SLSRA_LCM_20	Rock cliffs, ledges and shores	6
SLSRA_LCM_21	Scots pine woodland	9
SLSRA_LCM_22	Screes	1
SLSRA_LCM_23	Seasonally wet and wet grassland	5
SLSRA_LCM_24	Temperate montane scrub	3
SLSRA_LCM_25	Temperate shrub heathland	7
SLSRA_LCM_26	Valley mires, poor fens and transition mires	3
SLSRA_LCM_27	Windthrown woodland	1
SLSRA_LCM_28	Woodland fringes and clearings and tall forb stands	3
SLSRA_LCM_3	Arctic, alpine and subalpine scrub	2
SLSRA_LCM_4	Bare field or exposed soil	1
SLSRA_LCM_5	Base-rich fens and calcareous spring mires	5
SLSRA_LCM_6	Broadleaved deciduous woodland	9
SLSRA_LCM_7	Built-up area	3
SLSRA_LCM_8	Coastal dunes and sandy shore	10
SLSRA_LCM_9	Coastal shingle	6
SLSRA_NNR_2	National Nature Reserve (NNR)	7
SLSRA_NP_2	National Park	10
SLSRA_NR_2	Nature Reserve	6
SLSRA_RP_2	Regional Park	4
SLSRA_RSPB_2	Royal Society for the Protection of Birds (RSPB) Reserve	6
SLSRA_SAC_2	Special Area of Conservation (SAC)	6
SLSRA_SPA_2	Special Protection Area (SPA)	5
SLSRA_SSSI_2	Site of Special Scientific Interest (SSSI)	3
SLSRA_SWT_2	Scottish Wildlife Trust Reserve	3
SLSRA_WLA_2	Wild Land Areas	3
Water_Lakes_1	Pond	5
Water_Lakes_2	Lochan	3
Water_Lakes_3	Small Lochs	3
Water_Lakes_4	Medium Lochs	5
Water_Lakes_5	Large Lochs	7
Water_Lakes_6	Major Lochs	10
Water_Rivers_1	Minor river or tributary	4
Water_Rivers_2	Unnamed minor stream or tributary	1
Water_Rivers_3	Major river or tributary	8
Water_Rivers_4	Named minor stream or tributary	2
Water_Rivers_5	Lake	8
Water_Rivers_6	Tidal river or estuary	7
Water_Rivers_7	Canal	7

Computing Recreational Potential

We will now compare results in the ‘example’ region of interest (the Bush estate).

bbox <- get_example_bbox()

To compute Recreational Potential, as well as each individual component, we use the compute_potential function.

layers_hr <- compute_potential(hard_recreationalist, bbox)

Selected an area of 3.1e+07 m^2 .

layers_sr <- compute_potential(soft_recreationalist, bbox)

Selected an area of 3.1e+07 m^2 .

This returns a raster with 5 named layers.

1. SLSRA
2. FIPS_N
3. FIPS_I
4. Water
5. Recreational_Potential

To plot all layers together we could simply call terra::plot(layers). However, here we will separate the components plot from the RP plot.

#Set the colour palette for the heatmaps
palette <- terra::map.pal("bgyr")

First, the component layers.

Hard Recreationalist

terra::panel(terra::subset(layers_hr, 1:4), nc=2, col=palette)

Soft Recreationalist

terra::panel(terra::subset(layers_sr, 1:4), nc=2, col=palette)

Next, the Recreational Potential layers.

Hard Recreationalist

terra::plot(layers_hr$Recreational_Potential, col=palette)

Soft Recreationalist

terra::plot(layers_sr$Recreational_Potential, col=palette)

Defining a region of interest

A rectangular region of interest can be easily constructed using terra::ext by passing a vector of coordinates (xmin, xmax, ymin, ymax).³

xmin <- 300000
xmax <- 310000
ymin <- 700000
ymax <- 710000
bbox <- terra::ext(c(xmin, xmax, ymin, ymax))

Let’s visualise this area on a map of Scotland.

Scotland <- terra::vect(
  system.file(
    "extdata", "shapefiles", "Scotland", "Scotland.shp", package="biodt.recreation"
  )
)

terra::plot(Scotland)
terra::plot(bbox, border = "red", add = TRUE)

Editing a persona

Creating or editing personas is far easier using the Shiny app, some spreadsheet software, or even just a text editor. However, for completeness, we demonstrate how to edit a persona dynamically within R.

Let’s say we want to modify the Hard Recreationalist persona to better fit an off-road cyclist who does not want to cycle straight up a mountain. We will edit the slope scores and the cycle path scores.

# Note that assignment + edit in R automatically results in a copy!
custom_persona <- hard_recreationalist

custom_persona[["FIPS_I_NationalCycleNetwork_1"]] <- 5  # on-road, paved
custom_persona[["FIPS_I_NationalCycleNetwork_2"]] <- 10 # traffic-free, unpaved
custom_persona[["FIPS_I_NationalCycleNetwork_3"]] <- 10 # traffic-free, paved
custom_persona[["FIPS_I_NationalCycleNetwork_4"]] <- 10 # on-road, unpaved
custom_persona[["FIPS_I_RoadsTracks_5"]] <- 8  # access roads / tracks

custom_persona[["FIPS_N_Slope_1"]] <- 7  # no slope
custom_persona[["FIPS_N_Slope_2"]] <- 8  # gentle
custom_persona[["FIPS_N_Slope_3"]] <- 6  # medium
custom_persona[["FIPS_N_Slope_4"]] <- 4  # steep
custom_persona[["FIPS_N_Slope_5"]] <- 2  # very steep
custom_persona[["FIPS_N_Slope_6"]] <- 1  # extremely steep

We can calculate the RP values as before and compare the results.

layers_hr <- compute_potential(hard_recreationalist, bbox)

Selected an area of 1.0e+08 m^2 .

layers_custom <- compute_potential(custom_persona, bbox)

Selected an area of 1.0e+08 m^2 .

Hard Recreationalist

terra::panel(terra::subset(layers_hr, 1:4), nc=2, col=palette)

Custom Persona

terra::panel(terra::subset(layers_custom, 1:4), nc=2, col=palette)

Hard Recreationalist

terra::plot(layers_hr$Recreational_Potential, col=palette)

Custom Persona

terra::plot(layers_custom$Recreational_Potential, col=palette)

Interactive plots

To create a simple interactive plot is very easy using terra::plet.

terra::plet(layers_custom, 1:5, col=palette, tiles="Esri.WorldImagery", shared=TRUE)

You can explore different background tiles over at https://leaflet-extras.github.io/leaflet-providers/preview/.

Access to code & data

Code for the Recreational Potential model is available as an R Package hosted on GitHub at https://github.com/BioDT/uc-ces-recreation. These reports relate specifically to the version 1.0 release, which is also available on Zenodo at https://doi.org/10.5281/zenodo.15705544.

The model requires some input data that is too large to be bundled in with the package, but which can be downloaded from Dropbox after package installation using biodt.recreation::download_data(). This data is also included in the Zenodo listing.

The source for these documents and this website can be found at https://github.com/BioDT/ces-recreation-reports.

Funding

Funding for OpenNESS came from the European Community’s Seventh Framework Programme (FP7/2007-2013) under grant agreement no 308428, OpenNESS Project (Operationalisation of Natural Capital and Ecosystem Services: From Concepts to Real-world Applications.)

Funding for BioDT came from the European Union’s Horizon Europe Research and Innovation Programme under grant agreement No 101057437 (BioDT project, https://doi.org/10.3030/101057437). Views and opinions expressed are those of the author(s) only and do not necessarily reflect those of the European Union or the European Commission. Neither the European Union nor the European Commission can be held responsible for them.

Funding for SPEAK came from the Natural Environment Research Council – Growing Shoots Partnership and application co-creation bursary. NE/Y005805/1 _Growing Shoots.

CRediT statement

JMR: software; visualisation; writing. MT: software; visualisation; writing. CA: conceptualisation; data curation; funding acquisition; software. JD: conceptualisation; funding acquisition; project administration. SR: funding acquisition; software.

Acknowledgements

We wish to acknowledge significant contributions to a previous version of the Recreational Potential model by Will Bolton. JMR would also like to thank Tomáš Martinovič for assisting with understanding the requirements of BioDT.

We are very grateful to all the participants of the SPEAK project which this output is based on including: Alastair Leaver, Alice MacSporran, Brian Cassidy, Chris Pollock, David Giles, Jean Cowie, Joanna Gilliatt, Justyna Olszewska, Laura Taylor, Maddi Bunker, Rebecca MacLennan, S Mayes, Shaila Rao, Steve MacKinnon and Tom Gebbie, Active Stirling. We are equally grateful to an additional 18 participants who preferred to remain anonymous.

Correspondence

• {jand,chan}@ceh.ac.uk for enquiries relating to the BioDT or SPEAK projects, or ongoing and future work.

• chan@ceh.ac.uk for enquiries relating to the data sources and QGIS processing.

• {joemar,madtig}@ceh.ac.uk for enquiries relating to the code and anything else appearing in the technical supplement and/or developer’s report.

• simrol@ceh.ac.uk for enquiries relating to the biodiversity component of the BioDT project.

Footnotes

1. Since future updates to the code may be incompatible with this notebook, we will target the specific version released at the time this notebook was written.

2. This is where the UK Centre for Ecology & Hydrology is based.

3. A region of interest may also be given by a path to a shapefile.

Copyright

BioDT and the UK Centre for Ecology & Hydrology, 2025