A question about map scale for georeferencing

[WIECZOREK, John R.]

Susan, and all,

I agree that we might have benefited from having a separate field in which to record the map scale
during MaNIS georeferencing. That sort of information was supposed to be captured in the
DeterminationRef with enough detail that someone could find the same map. This was followed
inconsistently. I did a check. 98 of 158 different sources among your data included map scale.
Some of those that didn't were gazetteer sources. Still, I agree with you that we would have done
better if we made that requirement more clear. It certainly didn't help you with Colombia, where
none of the sources give the map scale. I looked through all of the rest of the Colombia records
and the scale of maps used ranges from 1:300,000 down to 1:500,000.

When we were doing MaNIS I didn't think it was appropriate to specify a minimum scale. The whole
uncertainty radius thing was meant to let us georeference no matter how good our source materials
were, and to capture the lack of specificity. I don't think that was a wrong decision, because
getting the points on the map was the most time-consuming thing. Now that they are mappable, I
believe that we will be able to do validation on our localities pretty well and fast. I think we
all have to realize that most localities can be refined, with smaller uncertainties, if we put
more work into it. I think the good news here is that we're in a better position now to improve
our data quality than we were before MaNIS.

Next July or so we are going to roll out the new BioGeomancer Workbench, which will consist not
only of automated georeferencing, but also of georeference validation and georeference editing
tools. My hope is that we can all use these tools to good effect when the need arises.  For more
about BioGeomancer, look at the new public web site released 1 Aug 2005
(http://www.biogeomancer.org).

John

At 09:09 AM 04 10 2005, you wrote:
I am wondering about the scale of maps that were generally used to complete georeferencing for
MaNIS.  It sure would have been nice if there had been a standard for referencing maps,
particularly that included scale.  When we did Ontario we used 1:50,000 maps leading to
georeferencing including extent calculations to be pretty much as good as they are going to get -
especially with the facilities offered by TouraTech software!  Inavailability of decent large
scale maps (even on paper) led to 'much more crude' georeferencing - extent and max errro
calculation specifically.  I never saw any expressed 'minimal effort' statement wrt map references
... so I am wondering how many georeferencers were dealing with say, atlas maps to georeference
... scale of 1,000,000 or even smaller?

The reason I ask is that we have someone from Colombia coming as an intern for a couple months and
it made sense to have her review our parsed data.  The maps we have mostly range from 1:400,000 -
1:500,000.  Of course this means that most extents will be determined as ½ to the next locality -
which was often done for our Guyana georeferencing.  My guess is that this is considered
acceptable - since otherwise things simply will not be georeferenced.

Hope all is going well with you.

Regards,
Susan