Authors: Jory Griffith and Nina Obiar
Review status: In review
The Protected Area Representativeness and Connectedness (PARC) indices measure the extent to which terrestrial protected areas, and other effective area-based conservation measures (OECMs), are ecologically representative, and well-connected - both to one another, and to other areas of intact natural ecosystems in the surrounding landscape (Hoskins et al., 2022).
The PARC-representativeness index provides a rigorous measure of the extent to which a system of terrestrial protected areas and OECMs is ecologically representative of the full range of environmental and biological diversity within any given spatial reporting unit (e.g. country, broad ecosystem time).
The PARC- connectedness index provides a rigorous measure of the extent to which protected areas and OECMs are functionally connected to one another and to other areas of intact natural ecosystems.
Together, the composite PARC indicator offers a measure of progress in the expansion of any system of protected areas and OECMs. The indicator is expressed in proportional (or percent) coverage with the position of any given reporting unit on this scale rigorously adjusted for the effects of both representativeness and connectedness.
This pipeline calculates a weighted geometric mean of the PARC indicator over a region of interest. The code to calculate the weighted mean was adapted from the "Calculating weighted geometric means of CSIRO BILBI indicator" script on the CSIRO data access portal
The PARC indices assess Goal A and Target 3 of the GBF, whether generated separately or as a composite, are used to monitor and report past-to-present trends in representativeness and connectedness by repeated calculation using best-available mapping of protected areas and OECMs at multiple points in time, e.g. for different years. They can also provide a foundation for assessing the contribution that potential additions to the system of protected areas and OECMs might make to improving present PARC scores, thereby providing a foundation for prioritising such actions.
There are no data or API keys required for this analysis.
BON in a Box contains a pipeline to calculate the PARC indicators for a given area of interest. The pipeline has the following user inputs:
Bounding Box and CRS: the user must select a bounding box and coordinate reference system (CRS) to be used for the analysis. This can be done by using the chooser to either select a country and/or region, or enter or draw a custom bounding box. Then, an appropriate CRS can be selected from the corresponding drop-down menu.
Start date: this input is optional. The user can select a start date for time series layers, in the format YYYY or YYYY-MM-DD. To perform the analysis on all available years, the user should leave this input blank. The layers start at 2020-01-01.
End date: this input is optional. The user can select an end date for the time series layers, in the format YYYY or YYYY-MM-DD. To perform the analysis on all available years, the user should leave this input blank. The layers end at 2024-12-31.
Temporal resolution: this input is optional. The user can select a temporal resolution for the query of STAC items by date, in the format ("P", time interval, and time unit, e.g. "P1Y" is yearly, "P1M" is monthly, and "P1D" is daily). If the temporal resolution is coarser than the temporal resolution of the time series, the layers will be aggregated with the aggregation method chosen below. The user should not put in a temporal resolution finer than that of the layer. The PARC dataset has a temporal resoltion of 5 years from 2000 until 2020 and then a layer for 2024. The user should leave this input blank if no start and end date was selected.
Spatial resolution: the user can select the spatial resolution of the rasters. This must be in the same units as the coordinate reference system (meters for projected reference systems and degrees for reference systems in lat-long). To use the original spatial resolution of the layers, the user should leave this input blank. If the spatial resolution is left blank, the CRS selected must be EPSG:4326.
Resampling method: the user must select a resampling method to be used when the analysis requires rescaling to a new spatial resolution and/or reprojecting of the raster layers. See gdalwarp for a description. This input will be ignored if there is no need for resampling.
Aggregation method: the user must select a method to aggregate items when layers are combined over time, e.g. if a courser temporal resolution is chosen. This input will be ignored if there is no need for aggregation (the temporal resolution of the original layers is selected).
This step returns the polygon for the country/region/area of interest. If a country/region was selected, it pulls the country/region polygon using the GeoBoundaries API, and outputs as a geopackage, projected in the crs of interest. If the user inputs a custom bounding box, it will return a polygon made from that bounding box.
This step first extracts the CSIRO Protected Area Representativeness and Connectedness (PARC) layers, then the CSIRO denominator layers, from collections on the GEO BON Spatio Temporal Asset Catalog. The layers are in EPSG:4326 and 30 arc-second resolution but the user can specify other coordinate reference systems and spatial resolutions. The CSIRO denominator layers are used to calculate the weighted geometric mean of the CSIRO PARC layers.
This step calculates the weighted arithmetic mean for the PARC layers to calculate the summary statistics over the area of interest.
Raster layers of indicator for each year: raster files of the PARC layers in geotiff format.
PARC summary: yearly weighted geometric mean of PARC in the area of interest. The value ranges from 0 to 1, where a higher value indicates that a higher percentage of the region's biodiversity and environmental variation is represented in protected areas. A low value suggests that some ecological types are underrepresented.
Time series plot: plot of the geometric mean of PARC over time in the area of interest.
Country: the country of interest, if selected.
Region: the region of interest, if selected.
Example output available soon.
Harwood, Tom; Ware, Chris; Hoskins, Andrew; Ferrier, Simon; Bush, Alex; Golebiewski, Maciej; Ota, Noboru; Perry, Justin; & Williams, Kristen (2022): PARC: Protected Area Representativeness Index: 30s global time series. v2. CSIRO. Data Collection. https://doi.org/10.25919/edwj-4b67
Hoskins, Andrew J.; Harwood, Thomas D.; Ware, Chris; Williams, Kristen J.; Perry, Justin J.; Ota, Noburu; Croft, Jim R.; Yeates, David K.; Jetz, Walter; Golebiewski, Maciej; Purvis, Andy; Ferrier, Simon. (2020). BILBI: Supporting global biodiversity assessment through high-resolution macroecological modelling. Environmental Modelling & Software 132:104806. doi: https://doi.org/10.1016/j.envsoft.2020.104806