Data analysis and modelling

We have extensive experience of data manipulation, analysis and modelling using GIS, statistical software (R, Minitab) and specialist modelling software (Smart-PLS, Spida, Gaffer).

We provide the following services:

  • Spatial data manipulation and processing in GIS:
    • River network processing, quality control and attribution.
    • Snapping and linking of point datasets on river networks (e.g. linking biological sampling sites with habitat data).
    • Derivation of network and catchment statistics for individual points on river networks.
    • Sampling strategy development and implementation.
  • Statistical analysis and modelling
    • Multivariate analysis of environmental data using PCA, Correspondance Analysis, Canonical Correlations, Genetic Algorithms and Neural networks.
    • Modelling complex causal relationships such as habitat and pressure and impacts relationships using Structural Equation Modelling and Partial Least Square Path modelling.
    • Modelling species/communities distribution using presence/absence or density data.
  • Spatial statistical analysis and modelling
    • Picture1Geostatistical modelling across landscapes and river networks.  Geostatistical models take account of spatial relationships and correlations to produce predictions across landscapes.  River networks are a specific case of geostatistical modelling as points are linked through a one-dimensional network (i.e. upstream-downstream relation). Models of species and habitat distribution across river network can be implemented using geostatistical techniques to take account of spatial interactions across landscapes and within networks.
    • Analysis of spatial river network data, development of models and implementation across entire river networks.  The example above shows a map of channel substrate size developed using geostatistics and River Habitat Survey data and implemented across 150,000 km of rivers in England and Wales (legend: red = coarse substrate; blue = fine substrate).