a collection of R and javascript tools for interactive mapping, ecology, and conservation

(05/08/2024)


Collection of spatial tools developed in r, leaflet and shiny that build upon leaflet draw functionality implemented in leafem and leaflet.extras that facilitate spatial planning and data visualisation. Note: these tools are functional examples for development and should be adapted if used (i.e. coded in a more efficient way).


Simple spatial tools

a) rotate polygons

Adds rotate functionality to leaflet by implementing the javascript leaflet plugin Leaflet.draw.rotate from https://github.com/onaci/Leaflet.draw.rotate in the leaflet_extras plugin Leaflet.Draw by using a custom version of the Leaflet.Path.Transform plugin. Useful for defining selecting non-gridded raster values for dynamic classification (see “Spatial decision tools” below for implementation).

b) draw plots

Creating fixed grid size plots (e.g. 100m x 100m, 1ha) within leaflet currently yet to be implemented. Note: leaflet uses EPSG:4326 coordinate reference system, so dimensions are approximate °0.00045 degrees lat/lon coordinates.



Dynamic raster classification


a) static rasters

D ynamic classification of static rasters by interactive reclassification of n bins and range for each interval. Adjusting the end point of the slider will dynamically shift the start point of the next bin. Example code below uses simulated raster data in R via gstat to approximate clustered raster cells.

b) dynamic rasters

Allows dynamic classification of rasters as above, but uses dynamic subsetting of spatial data from the current map extent. The code extracts satellite data (e.g. SST, Chl a) from the NOAA ERDAPP servers for the extent of the current map area (see below Interactive raster data visualisation for explanation), and then implements interactive reclassification of n bins and range for each interval. Note - the approach is similar to importing geoserver WMS tiles for a given map extent except uses raw data in .csv format to create raster files when zoom/recenter or changing classification from ordinal/continuous data.

Example uses Multiscale Ultrahigh Resolution (MUR) sea surface temperature (SST) from coastwatch at 0.01° resolution with a 7-day lag from present date.



Interactive spatial tools


i) Visualise raster time-series

Interactive visualisation of historical daily raster data from the NOAA ERDAPP servers.

Features:

  • Select satellite data sources from from ERDAPP servers

  • Specify dates for data

  • Use WMS base maps (coral atlas, ESRI imagery)

  • Allow custom palettes (e.g. NOAA degree heating weeks)

Example includes Great Barrier Reef and Ningloo (Western Australia) and data sourced from NOAA CRW and MUR. Note: currently uses re-rendering of raster images via .csv rather than direct conversion of .nc files, which is potentially inefficient - revisit with stars::read_stars over terra::rast at later stage



ii) Exclude map zones

Leaflet implementation of dynamic cropping of vector and raster layers to exclude zones from existing datasets prior to site selection. Users / stakeholders can draw zones/areas on the map, which are used to inversely mask the raster layer, updating the baselayers to exclude from further analysis. Code includes vector Allen coral atlas data and an aggregated raster layer generated by a Gaussian random field simulation.



iii) Dynamic spatial selection

Dynamic site selection tool for allowing users to interactively explore and analyze spatial layers through a leaflet map interface. Example includes simulated raster data of varying gridcell sizes (Temperature, Sedimentation, DHW, Wave Exposure, Chlorophyll, Salinity) and a “restoration score” (basic additive algorithm for categorical raster layers). Underlying code categorises each raster datalayer into ordinal values (see classify rasters) and calculates an unweighted mean, but can be implemented with any algorithm.