https://github.com/biodivhealth/sitetool

A tool for quantitative site selection.

https://github.com/biodivhealth/sitetool

Repository

A tool for quantitative site selection.

README.Rmd

---
output: github_document
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)
```

# {SiteTool}: A tool for quantitative field site selection




The Site Selection Tool is a R Shiny tool used to assess
the land cover characteristics of potential field sites.

The aim of this tool is to help researchers select sites in a
quantitative way for studies on ecology, public health, anthropology,
and more. Often, researchers aim to have field sites that exist along a
gradient of different land use types and characteristics. This tool
allows researchers to analyze the land cover characteristics of a list
of sites, and see where potential sites fall along a gradient of other
potential sites.

A ready-to-use browser version is available: https://nimirz.shinyapps.io/ss-analyzer/.

However, local installation is recommended following the below instructions.

## Installation

SiteTool is an R based tool. You first need an update-to-date installation of R (>= 4; https://www.r-project.org/). In an R interface or RStudio, please enter the following commands to install sitetool from [GitHub](https://github.com/):

``` r
install.packages("pak")
pak::pak("BioDivHealth/sitetool")

```
## Usage

Once the package and dependencies have installed, type the following to launch the app:

``` r
sitetool::run_app()
```

This will open an interactive interface which you can use for site selection. The tools is broken down into three main steps:

### Step 1: Select an area interest and type of land cover data

1.  Select the box icon located in the upper left-hand corner of the
    map. Use this to select your potential study site area. You may also
    manually enter the coordinates of your bounding box, or upload a
    shapefile. Shapefiles must be in GeoJSON format.
2.  Select an input raster data source. There are three default rasters (WorldCover, Elevation, and Human Footprint). You also have an option to upload your raster data.
    
    *Uploaded rasters should be in GeoTiff format. If your area encompasses multiple tiles, please merge before uploading. If you are using a categorical raster, make sure there is a color table associated with it for correct display. If you are using the browser version, please ensure you are using Google Chrome, as it is the only browser that supports GeoTiffs. (Note: the browser version can only support small raster files. If you need to upload larger files, please download the app from GitHub and run locally.)*

3.  Press **Add Raster**. The raster will be displayed on the map when finished.
4. You can continue adding rasters to cover all of your parameters of interests. Please use the map toggle to display the different rasters. 

### Step 2: Generate a list of sites

1.  Select a sampling procedure to generate a list of potential sites. You can select**random** sites or **village**       sites. Village sites will find all of the towns/villages/cities in the selected area
    using OpenStreetMap, and random sites will select sites based on
    simple random sampling. (*Note: if you select a very large area for
    **village** sites, the request may time out.*)
2.  For random sites, you must specify the number of sites required. You can additionally specify a distance between       sites. Major water bodies will be avoided, and points will be further from
    cities and roads than the distance provided. 
3.  You can additionally add some focal ("selected") sites. You can choose to select these on a map or upload a list of
    sites in CSV format. The CSV should contain a column labeled
    **site**, which has the site name, and columns with **latitude** and
    **longitude** in decimal degrees.
4.  Press **Go**. The sites according to your specifications will be
    displayed on the map.
5.  You have option to export the list of sites in this step.

### Step 3: Visualize results

1.  Input the distance from the center of each potential site you would
    like the land cover data analyzed from in meters. This distance is
    the distance from the point to each edge of the raster on all four
    sides, so a distances of 1000 meters (1km) would lead to an area of
    analysis of 4 km2.

2.  Press **Go**. The calculations may take a while to run. The plots
    and data will be displayed in the respective tabs when finished.

3.  The plot shows the values of your input sites
    relative to the other sites analyzed. Hover over each point to find
    the name of the site. For categorical rasters, the proportion of that land cover type found within the analysis distance is shown. For continuous rasters, the mean value of that measurement is shown within the analysis distance.
    
4.  You can add and remove sites from your "generated" and "selected lists. The map will also update.

5.  The stats tab indicates whether the selected sites are statistically different from the generated sites across land cover values. 

6. The data tables contain the results of the analysis for each site and can be exported. The following measurement values for categorical rasters are included:
    - **Proportion**: proportion covered by each land cover category
      within the analysis range
    - **Mean Patch Area**: the mean patch size for the land cover
      category in the area analyzed. A larger mean patch area indicates
      larger and more contiguous patches of that land cover type.
    - **Total Area (m2)**: area covered by each land cover
      category within the analysis range
      
For continuous rasters, the values calculated are mean, min, and max for the analysis area. 


## Input Land Cover Data

The app can support categorical land cover rasters such as: 

1.  [ESA WorldCover Viewer](https://viewer.esa-worldcover.org/worldcover/) (the default)
2.  [Dynamic World](https://dynamicworld.app/)
3.  [Copernicus Dynamic Land Cover](https://lcviewer.vito.be/)

Please visit the above sites to download a raster for your region of
interest. If your region covers multiple tiles, please merge tiles into
one file prior to uploading. Please ensure the raster CRS is WGS 84, and
the values for any land cover type are in one layer following the
numeric codes for the above products.

The app can also support rasters with continuous values, such as for
temperature, NDVI, or elevation. For these inputs, please ensure the
input raster has only one numeric layer.

It is recommended to use high-resolution products (\< 100m) due to the
scale of the analysis.

Owner

GitHub Events