In case you missed the 23 Sept Data Use Club practical session on making occurrence maps, you can visit the session page to watch the full recording and view the links to additional resources and documentation.
The 60-minute session featured Matthew Blisset, GBIF Software Developer, presenting on the using the GBIF API to download occurrences and map tiles to use in tools like QGIS and ArcGIS. The session also featured John Waller, GBIF Data Analyst, presenting on making maps using rgbif and making gridded maps using GBIF SQL downloads/species occurrence cubes.
Below, we’ve summarized some of the questions from the session and provided the pertinent links shared. We hope you use this thread to continue the discussion.
Examples
Lemur catta tutorial using rgbif
Links
Previous Data Use Club sessions related to this topic
Maps API documentation
rgbif maps
Creating maps from occurrences
Fetch maps of GBIF occurrences
SQL downloads
API SQL downloads technical documentation
Questions and Answers
Q
Has there been any plans/discussions on including the Equal Earth projection in GBIF?
A
We have now added this projection as the default for the quarterly analytics maps, for example for Active GBIF data publishing organizatiions.
Equal Earth projection may not be appropriate for occurrence data maps, since these need to be scrolled from side-to-side, which is likely not possible with the amount of data being shown. When a user is zoomed in slightly, even showing a large country, the size distortions of Mercator are much less of a concern, but maintaining the shape is useful. We will explore this further.
—
Q
Is it ok to use the GBIF plugin in QGIS for species occurrences? Or which one do you recommend?
A
Yes, you can read more in the GBIF technical documentation on Importing GBIF data in QGIS.
—
Q
I’m trying to use https://tile.gbif.org/4326/omt/{z}/{x}/{y}@1x.png?style=gbif-classic but it doesn’t work. What am I doing wrong? Is it the link that is badly formatted? In the browser it gives me a 400 error code. In QGIS it still says that the layer is unusable…How would I use (or how would the link be formatted) the add layer XYZ in QGIS?
A
If you just test it in a browser, you need to fill in the {z} {x} {y} yourself, e.g. with 0 0 0 for the whole world: https://tile.gbif.org/4326/omt/0/0/0@1x.png?style=gbif-classic
More detail is available in the GBIF technical documentation on Visualizations with the Maps API in QGIS.
—
Q
Do you have an available set of maps in ArcGIS for us to use?
A
Our existing web maps can be added to ArcGIS, or occurrence points added, or a GBIF download imported, all detailed in the GBIF technical documentation for Importing GBIF data into ArcGIS.
Other than these, we don’t plan to make additional map layers available in ArcGIS.
—
Q
In ArcGIS, I’m wondering if it is possible to increase the number of records to more than 100,000 in the future? Are you planning to add other parameters besides scientific name?
A
100,000 is a technical limitation and it’s unlikely we will increase.
The tool was launched in July of this year and it is currently only available on the desktop version of ArcGIS.
ESRI are extending ArcGIS Online to support GBIF occurrence data, including for queries returning more than 100,000 results.
Not all fields are available for occurrence points when using the ArcGIS geoprocessing tool. A limited number of fields are provided via the toolbox. Looking forward, we would like to incorporate more functionality into the toolbox and incorporate it into online ArcGIS, where it is currently unavailable.
—
Q
Are these queries via Python or other API logged for each dataset every time a map is created or connected?
A
No, they are not – just like someone browsing www.gbif.org doesn’t show logs against the dataset. If using ArcGIS and enabling the DOI, then the use is logged against the dataset.
—
Q
What is the difference between the rgbif and the gbifdb packages for data download? I’ve been using the gbifdb package and computing statistics on all biodiversity in France and it took 25 seconds with an NVMe SSD drive
A
rgbif is maintained by the GBIF Secretariat. rgbif interacts directly with data hosted by GBIF. SQL downloads through rgbif have the latest data, and the full range of columns, but as this is a shared service, there are some types of query that are not allowed.
gbifdb uses GBIF monthly snapshots we export to Amazon AWS, or one of these snapshots downloaded locally. It takes advantage of the Parquet columnar data format which allows some queries to be made very quickly — this is the same way we store occurrence data on the GBIF servers. For statistical queries it’s a good choice. Or, if you would like a local “copy” of GBIF to make various queries on, it’s a very useful tool for this. (You will need 225GB of disk space for a local copy, or less if you only need some of the data, e.g. one order, or a few countries.)
Querying using Google BigQuery is equivalent to using gbifdb, but (beyond the generous free allowance) the queries cost money to run.
—
Q
To use the rgbif map_fetch() function, can we also specify a licence (e.g CC0)?
A
Yes you can use map_fetch(license=”CC0”)
—
Q
Do SQL downloads also come with DOI, or can be structured to include DOI?
A
SQL Downloads do come with a DOI.
—
Q
Is there any way to filter occurrence records of native range of the species?
A
It depends if you have polygons for the native ranges of your species of interest. You can use the polygon to filter, either when retrieving the data from GBIF, or using ArcGIS/QGIS/rgbif/pygbif or other suitable tools.