Skip to Content

Resource hub

Habitat Suitability Model for Sabellaria spinulosa reefs in the UK: 2020 2020

Marine Seabed Habitats and Geology


This dataset was produced using an ensemble modelling approach, utilising the random forest algorithm to predict habitat suitability of Modiolus modiolus beds across the UK. Habitat suitability models require two types of input data, presence/absence data (also known as response variables) and environmental datasets (also known as predictor variables). The output of the model describes the probability of habitat occurrence as a percentage, with the overall resolution of the dataset being 300 x 300 m.

Another output shows the standard deviation of the predictive values of the habitat suitability model. This describes the variability of the habitat suitability models derived from fifty iterations using a random forests approach. The standard deviation can be used to assess the level of confidence, however this does not provide a full picture and should only be used as an indicator.

The predictor variables used within the model include:

  • Depth to seabed
  • Slope of seabed
  • Light attenuation coefficient of photosynthetic active radiation (Kd(PAR))
  • Kinetic energy at the seabed due to currents
  • Kinetic energy at the seabed due to waves
  • Seabed Substrate
  • Mean of annual temperature at the seabed (over 30-year period)
  • Absolute minimum of seasonal salinity

The response variables were extracted from:

  • OSPAR T&D Habitats Database (2018)
  • Natural England Evidence Base (2019)
  • Marine Recorder Database (2019)

More details about the input data can be found in the technical report:JNCC Report No. 718.

Resource type Dataset

Topic category Oceans

Reference date 2020·03·06

This dataset was produced using JNCC's in-house 'species distribution modelling framework' (JNCC, 2019), combining predictor variables (environmental datasets) and response variables (presence and absence datasets) to predict habitat suitability of Sabellaria spinulosa reefs in the UK. As true absences are particularly scarce in survey data, the presence of other habitats were used as a proxy for absences instead; this is referred to as pseudo-absence data.

Responsible organisation
Digital and Data Solutions, JNCC owner

Limitations on public access No limitations

Use constraints Available under the Open Government Licence 3.0

Metadata date 2022·08·01

Metadata point of contact
Digital and Data Solutions, JNCC

Temporal extent 1954·12·31 2020·03·06

Spatial extent
North 61.084677
South 48.973479
East 4.202898
West -12.969854
Back to top