latitude and longitude designation and collection data

[Gary Noonan]

        Several thoughts occur regarding the interesting discussions to date
about determining latitude and longitude.  One such thought is that I've
written a program (the RESEARCH INTERFACE) that partly needs the latitude
and longitude determination needs and want to update the program to meet all
such needs. At this stage I would greatly appreciate additional input from
my colleagues about their particular needs and requirements. Below is
information first about my background in working with GIS technology and
other computer matters. This is followed by information about the RESEARCH
INTERFACE program and a plea for feedback.
         I've worked with GIS software for several years, using both the DOS
and Windows version of Atlas GIS and published a review of the DOS version
in the June issue of Systematic Biology. Also I've extracted electronic
geography files from the federal Digital Chart of the World and have made
these available over the Internet.  The RESEARCH INTERFACE program that I
wrote  is,among other things, expressly designed to speed the determining of
latitude and longitude for records with locality data. I've been a member of
several task forces that worked on data standards and have also mailed out
proposed standards to entomologists and received their comments. Lastly, the
research monographs I write typically deal with tens of thousands of specimens.
         The RESEARCH INTERFACE program automatically determines latitude
and longitude for places that correspond to a database of about 99.99
percent of populated places in the U.S. When the program finds a unique
combination of geographic data that corresponds to a similar combination in
the database that accompanies it, the program automatically inserts a
decimal latitude and longitude into the appropriate research database
record. The program also inserts any missing political information such as
the county for a particular record. If only a state and county are given,
the program inserts a predetermined latitude and longitude for the county.
The program marks a radio button field to record the degree of accuracy of
the latitude and longitude. For example, 1 button notes at the degree
information is for county only as opposed to for a given locality. I intend
in future program versions to interface the RESEARCH INTERFACE with CD-ROMS
that will contain latitude and longitude for the entire world.
         The requisite data sets for the entire world will soon be available
and will be in the public domain.
         As noted by Al Newton the federal government has produced a GNIS
CD-ROM listing more than 2 million place and feature names for the United
States. This easy to use CD-ROM is preferable to most published gazetteers
and atlases. The current version already has many historical places that no
longer exist, and future versions will incorporate increasingly large
numbers of such historical places. The inclusion of such places is
particularly important when you are working with old specimens. Users can
search the CD-ROM, review the search results on the screen, print them out
or dump them into various files, including database files in DBF  format.
Its possible to search for and find localities that are only partially
spelled out on the specimen label. The federal data is public domain,
available for exporting into programs, for distributing to colleagues free
of charge, etc.
         Various foreign countries have produced electronic gazetteers.
However most such gazetteers are high priced. Data from the Canadian
government costs more than $700 for the country, and the user has to sign a
legal document promising member to provide the data to other people. This
makes it impossible to include it with  free programs such as the RESEARCH
INTERFACE. In fact since such foreign gazetteers would have to be merged
electronically into a single gazetteer for use with the RESEARCH INTERFACE
or other programs, the various foreign electronic data sets are of limited
use as regards interfacing with data management software.
         Many of you are probably familiar with the published set of
gazetteers for most countries that were published by the U.S. government and
are commonly called
"Board on geographic names " gazetteers or CIA gazetteers, etc.. These
printed gazetteers contain degrees and minutes (but no seconds) for foreign
countries.  The gazetteers also provide the major administrative unit in
which a feature occurs and identify a feature as a populated place,
mountain, stream, etc. The government is working on a prototype set of
CD-ROMS that will have the same information as the published gazetteers. The
prototype set will be available later this year but will run only under
Unix. I'm on the list of prototype reviewers and will post information to
Taxacom about my evaluation of the prototypes.
         Once the final set of foreign CD-ROMS is available, I want to take
data from them and from the federal GNIS US  database and combine it into a
set of CD-ROMS that will interface with the RESEARCH INTERFACE. Determining
latitude and longitude for all features will then be as easy as it currently
is for populated U.S. places with the RESEARCH INTERFACE. The latter program
will then serve as an easy to use database engine in which researchers can
determine latitude and longitude for sets of research specimens, and then
upload the information back to museums that lent the specimens. The program
will also be designed to import data from collections databases, determine
latitude and longitude and then export back to the main collections
databases--allowing for retrospective determinations when desired. It will
also have features for producing databases for use in GIS software and for
analyzing systematic data.
         I have thought about seeking NSF support for producing this
forthcoming version--some funds from somewhere will be needed. However,
discussions with NSF staff and various colleagues suggest that first there
must be a workshop in which systematists work together with GIS and
information specialists to determine the very best way (most economical,
most time saving, etc.) set of program standards for implementing the
forthcoming version. For example, much greater input is needed from
systematists about their needs. Much input is needed from GIS and other
computer specialists about the best ways to solve such needs. Such a
workshop would probably consist of approximately one third systematists and
two thirds computer workers. It would analyze needs and provide a blueprint
for implementing solutions. I then would seek NSF funding for the program
version with the worldwide latitude and longitude databases. The new version
would have many other features for data analysis.
        Because some people may not have seen the announcements about the
RESEARCH INTERFACE, I'll post an announcement separately. Separately so that
you can delete it if you've already seen it. A bit of honesty!! I just
discovered two bugs in the current version of the program and have fixed 1
and am working on the other. If you are thinking about downloading a free
copy, you might want to wait a week or so until I announced the fixed
version is available at the FTP site.
         Some questions on which I would appreciate feedback are as follows.
First, I know that there are people using the RESEARCH INTERFACE because
they contact me with questions. It would greatly help any future NSF
proposal for enhancing the program if users could let me know that they do
indeed use the program. Any such grant proposal would have to document
current users and document any desire by systematists for a program that
accesses databases for the entire world.
         It would also be interesting to know how accurate you feel your
data must be. My own experience in using GIS software is that typically I
make a map that covers all or a significant portion of a continent or at
least most of a major continental region. For such maps latitude and
longitude values for just county data are actually quite useful. Think about
how many square kilometers a dot actually covers on most maps, and you will
see that most often we don't need to worry whether a site 3 miles east of
Seattle is actually measured along a road, as the crow  flies, etc. Some
studies of smaller areas (such as ecological studies) of course make require
more precise data. For my own monograph type studies I've found that the
degrees and minutes in the printed federal gazetteers for foreign countries
provide sufficient accuracy for both regular maps in GIS and for thematic
maps that illustrate variation in morphological characters between
populations of species.
         How many people think that a program that encompasses determine
latitude and longitude on a worldwide basis would be useful? Would you
encourage development of such a program?
         What types of information should a program record to document the
degree of accuracy for a given record as regards latitude and longitude?
(The GIS workers who spend their entire careers working with GIS issues will
no doubt provide some good ideas in any workshop. ) Your ideas and needs
should be recorded now for use in any such possible workshop.
         What other issues are you concerned with?
          A NSF sponsored version of the RESEARCH INTERFACE can be
implemented on various platforms. The current version is written to run
under DOS and requires only a few megabytes of memory. However DOS is an
operating system on the way out--2 generations behind Windows 95. The
enhanced version could be written to run directly under Windows in FoxPro
for Windows version 2.6. However, Microsoft is considering not selling many
additional compilers for at version. The newest version of FoxPro is Visual
Foxpro. This version uses object oriented programming while previous
versions use the older xBase technology. Programming in object oriented
programming and maintaining programs written in object oriented techniques
are easier to do in the long run than are working with the older xBase
programing conventions. The various data screens created in Visual Foxpro
are extremely attractive from a graphics standpoint. Programs created in
Visual Foxpro can be more user friendly than those implementing under older
versions. The big problem however is that programs compiled under Visual
Foxpro require at least eight megabytes of computer RAM to run. Microsoft
assumes that most consumers will quickly move to more powerful computers
with such memory. Indeed, many computers sold to consumers for household use
have this much or more memory. However, systematists typically have very
small equipment budgets. A computer with 16 megabytes of RAM and a 75
megahertz Pentium processor costs about $2200 from vendors such as Dell. Is
this too steep a hardware requirement for systematists?
         If you are using the RESEARCH INTERFACE, what do you like and what
do you dislike. Please be candid.
        The program uses latitude and longitude rather than other units such
as  UTM units. The problem as I understand it for the latter units is that
they don't match up when countries are compared on a global basis.--I've
never used UTM myself because GIS programs mostly work with latitude and
longitude--Any comments on the accuracy of UTM units when used globally.
         The RESEARCH INTERFACE currently uses decimal latitude and
longitude because the DOS version of Atlas GIS preferred decimal to degrees
minutes and seconds. I've often quite used to decimal degrees and find them
conceptually easier to visualize than minutes and seconds with their darn 60
units for minutes and 3600  units for seconds. Any comments about ease of
use regarding decimal reverses the old fashioned minutes and seconds.
(Windows version of Atlas GIS can use both methods. )
         I would like the forthcoming version of the RESEARCH INTERFACE to
be constructed so as to handle data for many different systematics
disciplines. I know about the data standards for entomologists but need
information about standards for other disciplines. If you are in a different
discipline that has established standards, could you please send me a copy
or let me know where I might obtain one.
         Well, that's my questions for the moment. I'll appreciate any
feedback you may care to give. Particularly interesting will be feedback
about the desirability of what I am thinking of doing with the future
version of the RESEARCH INTERFACE. Do you want such a version? A successful
proposal to NSF would have to document that systematists desire the software
tool in question. Also any suggestions for a more catchy program name will
be greatly appreciated. Thanks for your patience in wading through this message.



  *************************************************************
  * Gary Noonan, Curator of Insects, Milwaukee Public Museum  *
  * 800 W. Wells,  Milwaukee, Wisconsin 53233 USA             *
  * and Adjunct Associate Professor of Zoology, University of *
  * Wisconsin-Milwaukee carabid at mpm1.mpm.edu                  *
  * voice (414) 278-2762  fax (414) 223-1396                  *
  *************************************************************