[UW-GIS-L] Q: joining in Arcview 8.3
hgreen at u.washington.edu
Wed Aug 17 10:52:26 PDT 2005
Presuming that you do not yet know the second nearest station to each
birthplace, I would write an aml that would loop through each
birthplace. It would:
a) put one birthplace point into a special temporary coverage.
b) Use NEAR to establish ID and distance of pollution stations.
c) Put this information into one table and sort.
d) Transfer the top three records into three pairs of fields which you
have created in the birthplace coverage.
Then you could:
1) join birthplaces with the closest station.
2) reselect the unmatched records.
3) join these with the second closest station. ...
A completely different approach would be to build an interpolated TIN or
lattice for each pollution category for each time step, and use TINSPOT
or LATTICESPOT to assign an interpolated value to each birthplace point.
Another approach is to seek help from a meteorologists (who probably
don't read this list) about using spatial/temporal interpolation to
reconstruct missing station data.
Constructing a suite of TINS for each time step will chew up your CPU,
but would not be that hard to implement. I think interpolated values
will be superior to nearest neighbor values.
I assisted a cardiac study that modeled particulate concentrations as a
function of slope and uphill area. That was a rather bold assumption,
but it might be interesting (depending on your resources) to explore
constraining that model (or another one) to your data points.
If you pick an approach, I can help you flesh it out.
Carole B. Rudra wrote:
> Hi all,
> I'm doing a project in which I'm joining birth records to air
> pollutant monitoring sites. Because not all sites have data at all
> times, I'd like to join records to the nearest site, then the
> second-nearest, and then perhaps the third-nearest. Is there an easy
> way to do this in ArcView using the join procedure?
> Carole Butler Rudra, MPH
> Doctoral Candidate
> UW Dept of Epidemiology
> Phone 206-215-3050
> Fax 206-386-3173
> clb3 at u.washington.edu
> Uw-gis-l mailing list
> Uw-gis-l at u.washington.edu
hgreen at u.washington.edu 511A Condon Hall 206-685-7981
University of Washington
Department of Earth and Space Sciences
Condon Hall (Box 351310)
Seattle WA 98195-1310
More information about the Uw-gis-l