Clean reference data LBBG_IJMUIDEN

[sannegovaert]

When we download the reference data from UvaBiTS, the data is automatically mapped to the Movebank data format.
Here is the reference data for LBBG_IJMUIDEN in Movebank format.

Questions for @jshamoun and @sshinne

1. ses.start_date = deploy-on-timestamp and ses.end_date = deploy-off-timestamp
deploy-off-timestamp is now always set at midnight.

   • Can you confirm that the deploy-on-timestamp and deploy-off-timestamp is correct for alle deployments?

The deployment thus only starts when a bird is tagged and released. And the deployment ends when the bird has died or the tag has been removed. deploy-off-timestamp is now always set at midnight. We do not want to publish measurements when the bird is in captivity (eg rehabilisation center) or the tag in somebody's car or office :)

2. animal-sex
   The sex of many of the birds is unknown.

   • Please update animal-sex if more information is available.

3. ind.remarks = animal-comments and ses.remarks = deployment-comments
   This needs to be cleaned up.

   • 3.1 There are measurements in animal-comments that need to formatted so they are mapped to deploy-on-measurements. They need to be formatted in key:value | key:value pairs.
   • 3.2 Some of the text in ind.remarks = animal-comments is exactly the same to the text inses.remarks = deployment-comments. Please make sure that this is not duplicated.

4. Be aware that deployment-end-type is filled in based on the presence of 3 words in session remarks. Other values from Movebank vocabulary cannot be consistently derived.

deployment-end-type set to	Session remarks contains the word
dead	dead
equipment failure	defect
equipment failure	malfunction

[sshinne]

1. deploy-on-timestamp is always correct for all deployments. Deploy-off-timestamp is correct, but, when deployment starts, typically it is set to a moment in the future. Once it becomes inactive for any reason, the deploy-off-timestamp is changed. For this project it looks like deploy-off-timestamp is NOT always set at midnight.

2. We do not have more information on this.

3.1 Measurements in animal-comments should not be mapped to deploy-on-measurements. This information should not be included in the uploaded data.

3.2 Similar to Issue #260: animal-comments and deployments-comments should only be used to populate other fields when that makes sense (e.g., to set a value in deployment-end-type), but otherwise these comments should not be passed to ind.remarks or ses.remarks. We need to leave this as is in our database, and they should not be uploaded to Movebank to remarks fields.

4. We are willing to look into this, but we are wondering if you have any suggestions for language we should use to replace it. We add "dead logger" into the database when we have no GPS tracking fixes coming in for years but we have not recovered the logger or the bird. Track sessions can end because we find a dead bird, the tracker has been removed from a live bird, the tag malfunctions or the tag falls off (i.e. data is still being collected but in one location for a long time).

[sshinne]

Also, please note that any individuals with “chick” or “juvenile” in animal-comments or deployment-comments should not be included in the uploaded dataset.