CVS embargoes¶

1 critical issue; 1 non-critical issue

critical issues¶

1. remove plot/plot_.locationNarrative (directions_to_place) from those rows that have had coordinate fuzzing
   • plot.confidentialityStatus might say which rows these are, or otherwise this would require adding back plot.realLatitude/realLongitude in order to determine which rows these are

non-critical issues¶

1. remove embargoed columns from the staging tables
   • embargoed rows have been omitted from everything, but columns have only been omitted from the normalized DB

e-mails¶

_from Mike Lee:_

I omitted the embargoed data with any complicated options. I included data that had simple "fuzzing" of geocoordinates where the actual lat/long are in the reallatitude and reallongitude and the public reduced-precision coordinates are in latitude and longitude.

If the data are to be publicly accessed, something I think was not an option when we gave you the data, only the latitude and longitude fields should be used.

There are two kinds of embargoes. The first is user-specific, something that is stored in the embargo table and was never developed, so that's why the table is empty. The second is the plot.confidentialityStatus field. If that is >=4, then it's embargoed from the public with something other that geocoordinate fuzzing. These have been excluded.

Does this mean that we cannot redistribute CVS, or can we redistribute
it if the private fields/embargoed plots are removed?

Correct, it should not be redistributed in its current form.

So if we delete the values in realLatitude/realLongitude, this is
sufficient for the data to be safe to make public?

Yes, if those values are removed. The authorE and authorN should also be removed, those could betray the real location as well. We don't expose those in VegBank's interface.

_from Bob:_

We need to remove the private fields to distribute CVS

Below is what I think we need to do. I cc Michael for verification

In the source data, consider the following fields
plotData: confidentialityStatus (0=Public, 1=block all; 2= county only)
plotData: realLatitude OMIT
plotData: realLongitude OMIT
plotData: locationAccuracy RETAIN (except OMIT if confidentiality status = 2)
plotData: publicLatitute RETAIN
plotData: publicLongitude RETAIN
plotData: realUTME OMIT
plotData: realUTMN OMIT
plotData: realUTMZone OMIT
plotData: authorCoordinateType OMIT
plotData: authorCoordinateUnits OMIT
plotData: authorE OMIT
plotData: author OMIT
plotData: authorZone OMIT
plotData: authorDatum OMIT

Does OMIT mean the data is sensitive, or just that we don't need to include it in VegBIEN?

Sensitive. Expunge it from BIEN. But, wait for Michael to verify

_from Mike Lee:_

I agree with Bob except three things.

1) location accuracy can be retained as far as I can tell. It just says how accurate the base-coordinates are. If they are not very good, then the fuzzed coordinates might be even worse than one suspects.

2) plot.locationNarrative should be Omitted - this could tell you exactly how to get to the plot.

3) plot.confidentialityReason should be omitted.

I note that the 3rd to the last line should read:

plotData: authorN OMIT

The N was omitted.

Other than that I think we are good.