Skip to content

Spatial discretisation methodology vignette

Starting point, layers involved in modelling:

  • RPG: give the crop at plot scale
  • Soil depth (currently BDGSF but soon Info&Sols at 90m)
  • Soil AWC (Info&Sols at 90 m)
  • Soil texture (optional)
  • Climate (SIM2 or DRIAS/Explore2 at 8km)
  • Zones of interest for the manager

Methodology:

  • starting point: extent of zones of interest
  • Aggregation of AWC raster into classes
  • Aggregation of soil depth (if raster based)
  • Intersection with soil depth (see how to proceed...)
  • Intersection with soil textures
  • Affectation of mean climate data on each polygon
  • You are ready to run the model!

Aggregation of raster data

Define a categorization for transforming raster into polygon of homogeneous values. Number of class has to be decided in order to get a trade off on number of polygons.

How to proceed intersections (See #21)

Compute mean and standard deviation of data to intersect and choose by polygon if a cut is necessary based on standard deviation threshold. The threshold could be define on the class magnitude but these latter are not homogeneous...

TODO: Write vignette implementing this methodology