# dots on maps

[Gerald Noonan]

Hugh Wilson and other colleagues have produced some very interesting comments
about handling geographic distribution data. Below are some thoughts regarding
such data.

At 08:56 AM 11/4/99 -0500, Hugh Wilson wrote:
  Creation and implementation of look-up tables or
>other devices to produce coordinate points that are not present on
>the label consumes time and and resources and the result often
>carries a significant element of potential error.

        Obtaining latitude and longitude for specimens by using printed
gazetteers and maps can indeed be very time-consuming. Several years ago I
decided to move all the geographic information I process during research into
GIS format and to use GIS software to analyze the biogeography of the insects I
study. After spending time trying to determine coordinates from gazetteers and
maps, I wrote a program called OmniMap. The currently available version of the
program lets the user enter label data and speeds such entry by options such as
allowing the user to have the program automatically enter into each new record
repetitive data -- for example the name of the state, county, etc. I obtained a
federal database of populated places in the United States and converted the
latitude and longitude into the decimal form used by GIS programs.
        The program greatly sped up the operation of entering label data and
finding latitude and longitude coordinates for same for sites that were
populated places in the United States. After the user enters information for a
populated site, the program looks up the latitude and longitude for that site
and inserts it into the current record.
Other sites still had to be looked up in gazetteers and on maps. However, even
for such sites, the program has some options for speeding up data entry -- for
example it has a window for converting regular degrees into decimal format and
inserting the decimal degrees into the current record.
        Recently I took the public domain data on a federal CD-ROM and
converted
it into a database for use with OmniMap. This database has approximately 1.1
million records. It contains all the sites found on U.S. topographic maps as
well as a number of historic sites that no longer exist. Students have been
using this larger database for a few weeks and find that obtaining latitude and
longitude for U.S. sites is a relatively rapid and easy process.
        Even more recently I obtained from the Defense Department a CD-ROM with
a tab delimited file of more than 5 million foreign place names. This has been
converted to a dBase compatible database with latitude and longitude for each
place. I still have some work to do as regards writing code for having OmniMap
take latitude and longitude values from this database. I probably will not get
around to writing this code until early next year. There also are a few bugs
regarding the U.S. database of 1.1 million records. When time permits, the
program code will be updated and all databases put on the World Wide Web for
use by colleagues -- probably first quarter of next year.
        The limited version of OmniMap is available for anyone who wants. This
is the version that uses the small database of populated places in the United
States. It can be downloaded from
http://www.mpm.edu/collect/invert/omnimap.html

        OmniMap was written several years ago in the DOS version of FoxPro. To
my happy surprise the DOS version is able to handle the quite larger databases
described above. -- the reason I don't simply put the large 1.1 million record
U.S. database online is that I'm busy writing a grant proposal for the rest of
the month and then will be away all of December. Minor changes in the program
code are needed to fix some bugs that occasionally interrupt operation with the
larger database. The foreign database will require some further massaging
because of issues such as handling characters with accents or diacritical
marks. The OmniMap program version that will be made available next year will
also contain a number of improvements designed to speed data entry.
        When the various larger databases are fully operational with OmniMap,
they along with a new version of the program will be available for downloading
from the World Wide Web.

>Most users (or potential users) on web-based floristic products are
>*not* academic or scientific specalists.  The rationale for mapping
>should involve the production of visualizations that will answer
>users questions, which are usually framed within a geographic context
>defined by political boundaries of some sort. .

        Systematists should begin thinking about converting all their research
geographic data into GIS format. Once the data are in such a format, GIS
software can be used to quantify the biogeographic relationships between
species and to do analyses that formerly were not possible. The databases can
also be put online for downloading by colleagues. GIS compatible databases can
be combined and partitioned as desired so that workers can do comparative
biogeographic studies.
        The databases can also be made available online with a variety of
Internet GIS program servers that show World Wide Web users a map with symbols
on it. The user can click on a symbol and have the GIS server program display
the records on which the symbol is based. If desirable, analyses can even be
done on the data displayed on the World Wide Web. (I hope to have my museum
purchase such a server program under our K-12 license next year and will
gradually put maps online.)

>"The recent movement of mainframe-based GIS systems to the
>pc and associated software that allows web-based mapping products to
>be generated from these systems .... will probably produce a different future
working environment "

        These programs can be surprisingly inexpensive. A few years ago I
arranged for my museum to obtain a site license for ArcView. The license for
this GIS program was under the K-12 program of Environmental Systems Research
Institute and cost less than $400 for the entire museum. It lets us run as many
copies of the program as desired. And the license includes the use of several
thousand dollars worth of GIS data in the form of maps at many different
scales, topographic data and contour lines, and ArcAtlas. The latter Atlas is a
CD-ROM with very interesting GIS maps of climate zones, vegetation zones, zones
based on soil type and amount of solar radiation, etc. The ArcAtlas allows
biologists to examine distributions relative to a variety of factors such mean
annual precipitation, temperatures, etc.
        ArcView was originally a program designed for viewing data produced
with
ArcInfo--a program the typically requires a dedicated workstation. However,
recent versions of ArcView have incorporated many features for analyzing and
generating data. And ESRI has produced a variety of extensions which can be
purchased to extend the analytical capabilities of ArcView. The extensions are
very cheap for educational institutions.
        I've found that ArcView with selected extensions provides a wealth of
tools for analyzing biogeographic relationships. A manuscript is in press with
the Journal of Biogeography showing some of the types of analyses that may be
done. The interesting thing about ArcView is that while it can do very
sophisticated analyses it can also be very easily used for simpler tasks such
as making distribution maps. Typically a student can use the tutorial for the
program and learn how to make maps and do the simpler types of analyses in just
a few hours.