Landscape Patterns
How to discover and to describe
Landscape Metrics
- Problem no. 1: which of the hundreds of spatial metrics should we choose?
- Problem no. 2: many landscape metrics are highly correlated…
![]()
Possible Approach - PCA of type samples of landscape metrics
We performed a principal component analysis (PCA) using 17 landscape-level metrics:
![]()
PCA of landscape metrics
- First two principal components explained ~71% of variability
PCA of landscapes
The result allows to distinguish between:
- simple and complex rasters (left<->right)
- agmented and consolidated rasters (bottom<->top)
However, there are still some problems here…
PCA of landscapes
- We performed a second PCA using data from the United Kingdom only
- Next, we predicted the results on the data for the whole Europe
PCA of landscapes
Issues with the PCA approach:
- Each new dataset requires recalculation of both, landscape metrics and principal components analysis (PCA)
- Highly correlated landscape metrics are used
- PCA results interpretation is not straightforward
IT metrics
- Five information theory metrics based on a co-occurrence matrix exist (Nowosad and Stepinski, 2019, https://doi.org/10.1007/s10980-019-00830-x)
- Marginal entropy [H(x)] - diversity (composition) of spatial categories - from monothematic patterns to multithematic patterns
- Relative mutual information [U] - clumpiness (configuration) of spatial categories from fragmented patterns to consolidated patterns)
- H(x) and U are uncorrelated
IT metrics
2D parametrization of categorical rasters’ configurations based on two weakly correlated IT metrics groups similar patterns into distinct regions of the parameters space
![]()
Exercises
- The marginal entropy and relative mutual information can be calculated using the landscapemetrics package’s functions:
lsm_l_ent()
and lsm_l_relmutinf()
. Calculate both of these metrics for the exdata/lc_small.tif
raster.
- Read the
exdata/lc_europe.tif
raster using rast() from the **terra** package and the
exdata/polygons.gpkgvector data using the
read_sf()function from the **sf** package. Calculate the marginal entropy and relative mutual information for each polygon using the
sample_lsm()` function.
- Join the calculated values with the polygons (see https://r-spatialecology.github.io/landscapemetrics/articles/irregular_areas.html for more details).
- Calculate SHDI and AI for the polygons. Compare the values of SHDI and AI with the marginal entropy and relative mutual information (e.g., using a scatterplot or by calculating the correlation coefficient). Are the results similar?
- (Extra) Create your own polygonal grid using
st_make_grid()
function from the sf package for the area from the exdata/polygons.gpkg
file. Calculate the marginal entropy and relative mutual information for each square using the sample_lsm() function. Visualize the results.
IT metrics
These metrics still leave some questions open…
- Relative mutual information is a result of dividing mutual information by entropy. What to do when the entropy is zero?
- How to incorporate the meaning of categories into the analysis?
PCA 1
PCA 2
PCA 2
PCA 1
PCA 2
PCA 1 PCA 2 UK EU
Entropy
Relative mutual information
Land cover data
Parametrization of two IT metrics
Parametrization of two IT metrics