Where to describe equipment and protocol used for machine observations

[peterdesmet]

Original question by @sarahcd:

Here I compiled the available sensor information (see the Movebank-format “reference data”) under samplingProtocol in the event and occurrence tables but that seemed a bit of a messy way to describe equipment, especially for cases where more sensor info is available and where detection limits or accuracy might be critical to interpreting measurements. For example, Movebank has a number of attributes used to describe tags and tag deployments on animals. In my experience these are a pretty basic set of attributes that can cover most information commonly available for most animal tracking data. Additional DwC terms could also take advantage of more thorough vocabularies, e.g. the SensorML vocabularies the BODC maintains for marine sensor data.

Answer by @albenson-usgs:

Hopefully someone will be able to give a more nuanced answer to this but I think the decision was to put all of that in the extended measurement or fact file. I noticed there were several pieces of data that were in your original version and didn’t make it into the DwC-A version. I think things like tag manufacturer, tag model, tag tech spec, sensor type, attachment type, duty cycle, etc could all go into extended measurement or fact.

Current example of samplingProtocol:

Lotek GPS collar deployed on free-ranging animal. Scheduled to collect location fixes every 5 min.

[peterdesmet]

I agree with @albenson-usgs, that info could be added as facts about the deployment:

id	measurementType	measurementValue
IDCoy_P3_1:deployment1	attachment type	collar
IDCoy_P3_1:deployment1	deployment end type	removal
IDCoy_P3_1:deployment1	deployment id	166760_Winter
IDCoy_P3_1:deployment1	duty cycle	5-min fixes
IDCoy_P3_1:deployment1	manipulation type	none
IDCoy_P3_1:deployment1	study site	Idaho National Engineering Laboratory
IDCoy_P3_1:deployment1	tag manufacturer name	Lotek
IDCoy_P3_1:deployment1	tag model
IDCoy_P3_1:deployment1	tag readout method	tag-retrieval

I think all of these are related to the deployment and not to the capture/recapture. Ideally they would also have a measurementValueID that links out to a definition for each of these facts/settings.

[sarahcd]

Thanks @albenson-usgs and @peterdesmet. Suggested additions to the MOF table made for V3 of the Mahoney case study, including measurementValueID.