In raff-k/VLSM: Vector-based Landscape Metrics

knitr::opts_chunk$set(
  fig.align = "center",
  collapse = TRUE,
  comment = "#>"
)
# knitr::opts_chunk$set(fig.pos = '!h')

knitr::opts_chunk$set(fig.pos = 'H')

1 Vignette Info

Most common landscape metrics are calculated on a raster-basis. However, sometimes landscape information come along in vector-formats. For the calculation of landscape metrics, one must perform a vector-to-raster-transformation often resulting in a loss of information. In VLSM we implemented various landscape metrics based on vector data in R. However, sometimes it is easier to process data with only few clicks using a desktop GIS. Therefore, we created compatible R scripts being accessible in the open-source desktop GIS QGIS. In the following, a quick manual on how to use these functions in QGIS. If you want to use these functions via R, please refer to here.

2 Packages and Software Utilities

The open-source desktop GIS QGIS enabele the possibility to integrate R-scripts. However, QGIS does not except the standard .R-format, instead we developped R-scripts which are compartible to QGIS and can easily be loaded in with only few clicks. You can find the VLSM R-scripts for QGIS here.

We tested our implementation on Windows with the following software and packages: Before integrating R into QGIS, one must have installed R (download-page). Our implementation was tested using R version 3.5.2 (2018-12-20) -- "Eggshell Igloo". Furthermore, we recommend to install the necessairy packages sf, sp, raster, link2GI, rgrass7, RSAGA, dplyr, and of course VLSM via R. QGIS 3.6 Noosa: We recommend QGIS> 3.x which can easily be installed by the OSGeo-installer. In QGIS one must enable the option for R-processing, first. An comprehensive instruction can be found here, or here. It may be that R is not visible or included in the setting menu. If that is the case the Processing R Provider must installed as plug-in (see here for help).

Note: One can set R's library directory in QGIS. Installing necessairy packages may be better done inside R.

When the setup of the setting was successfull, the QGIS desktop should look similiar to the following screenshot:

# LOAD REQUIRED LIBRARIES
pacman::p_load(VLSM, magick)

# LOAD THE DATA OF THE PACKAGE:
data(VLSM_QGIS, package = "VLSM")  

image_Intro <- magick::image_read(VLSM_QGIS$image_intro)
image_Mesh <- magick::image_read(VLSM_QGIS$image_mesh)
image_Urban_Sprawl <- magick::image_read(VLSM_QGIS$image_ffe)

# ![](sf_xfig.png)

```{R, messages = FALSE, invisible = TRUE, echo=FALSE, fig.cap="QGIS setting to access R functions", out.width = '100%'} plot(image_Intro)

# 3 Calculation of Landscape Metrics

We use Corine Land Cover (CLC) data, which are in detailed explained in the [R-vignette](https://github.com/raff-k/VLSM/blob/master/vignettes/vignette_R.pdf) of this package.

## 3.1 Example 1: Effective Mesh Size

The *effective mesh size* is calculated with the tool _Mesh_. The output delivers a table with the corresponding mesh sizes [ha].

```r 
plot(image_Mesh)

3.2 Example 2: Urban Sprawl

The Urban Sprawl is calculated with the tool Urban Sprawl. The output delivers a table with the corresponding degree of urban sprawl [%], and a geometry showing the grid lines, which were used for calculation.