gbif data and annotation

[xhdong-umd]

@chfleming @jmcalabrese I checked out some details about gbif and rgbif. Looks like

• we can use rgbif to download occurrence data (this require a gbif account, and some condition/filtering using rgbif)
• If we had some data with location and times, we can try to annotate them with Movebank EnVData service.

I haven't really download some data yet, is this what you you have in mind for this direction?

rgbif looks to be a powerful and flexible package, how should we design the workflow with the annotation step?

My first impression is I can write some function to extract bounding box and time range from the downloaded data, then it'll use the same function to request annotation.

The first steps are to build some scripts into the package. Then with some usage we can have some ideas about possible web app structure, and I can start to design and implement a web app if needed.

[xhdong-umd]

I downloaded some example data. We can set a condition to only download observations with location. I got a 3849 rows, 250 columns data.frame, with time range from 1770 to 2020.

I guess this kind of occurrence data often have a long time span.

• Since the MODIS satellite data only started from 2000, we may need to first filter by satellite data range. This reduce to 2543 rows.
• Though this is still too much. How should we request annotation for this kind sparse occurrence data?

[chfleming]

That all sounds reasonable to me. You are filtering by species, right?