GeoWeb: Probably the first such interface to hit the web. Brandon Plewe wrote the map interface while working and finishing his Ph.D. at SUNY/Buffalo.

In 1994 Java was still called OAK and still being developed for toasters, NASA's EOSDIS project was finishing up an X client to serve as their geographic database search interface, NOAA was experimenting with static gif images of maps, and Brandon Plewe was writing the first interactive geographic area selection interface for the WWW.  Brandon used the Xerox PARC map server to generate his maps and consequently was restricted to the projections the Xerox PARC server offered. Those are: the Equatorial Cylindrical Equidistant projection, the Sinusoidal projection, and the Mollweide projection.  For some reason Brandon chose to use the Sinusoidal, probably because of the three choices that's the projection that least distorts the U.S.

ADL:  The Alexandria Digital Library is head quartered at UC-Santa Barbara but has subgroups at CU-Boulder and SUNY/Buffalo.  The map interface for the ADL appears to have some heritage in GeoWeb in that they too use the Xerox PARC map server.  ADL, however, lets the user choose any of the three projections offered by Xerox PARC.  Unfortunately the three offerings are all "world view" projections and as such are fairly inadequate for the polar community.  The ADL interface is a prototype, and still in development, so to check it out you have to sign up to be a beta tester and get yourself a password and whatnot.  But here's a screen shot.

WebGLIS: The web version of the Global Land Information System is maintained by the EROS Data Center (EDC) in their capacity as part of the USGS.  EDC is also a Distributed Active Archive Center (DAAC) in NASA's Earth Observing System - Data Inventory System (EOS-DIS) project and so have also been involved in the development of the IMS mentioned below.  It is unclear to me if there are any links between GLIS and the IMS apart from EDC's involvement in both.  The development of X-GLIS and later development of WebGLIS seems to have paralleled development of the V0IMS X-client and the later development of the WWWGateway.  I suspect the former were developed first.

WebGLIS is notable in that it does not let users select an area on the Equatorial Cylindrical Equidistant projection which is normally the only, or at least the default, projection used in these interfaces.  Indeed WebGLIS does not offer any "world view" projections at all.  The only projection in WebGLIS is an Orthographic projection which can be rotated and zoomed.  The map itself is a gif image that gets re-generated on the fly every time the users changes something.  And the only area selections allowed are ranges which means that even though the interface uses an Orthographic projection the server is using the Equatorial Cylindrical Equidistant projection.  This also means it is terribly difficult to include the pole in the selected area.

V0IMS:  Strictly speaking the Version 0 Information Management System isn't a web interface.  It's the prototype X client for NASA's EOS-DIS project.  But it is a geographic database search interface, and it's the first such publicly accessible interface I've worked with, and it is on the web.

It was because of my involvement with the V0IMS, and my employment at the National Snow and Ice Data Center (another DAAC as well as the World Data Center A for Glaciology), that I first started to notice the problems with most of these interfaces and first started trying to champion the interests of the polar community.  My efforts to explain the problems were generally met with blank stares and eventually resulted in the paper mentioned above.  A talk I gave at the V0IMS developers conference in April 1997 - which details specific problems with the X client and contains several screen shots - is also online.

I have also put the V0IMS X client itself online as an experiment with Broadway (X11R6.3).  To access it via the web you need an X server on your machine and the Broadway plugin (libxrx) for your browser.  If you don't want to bother with installing the plugin, and you have an X server on your machine, you can still access it via telnet.

WWWGateway:  Because access to the X client is restricted to people with X servers the WWWGateway project was conceived to allow access to the same data via the world wide web.  Because Java was still a fledgling technology at the time, and the JEST interface (see below) had already begun development, the decision was made to try to implement an HTML only interface.

In early versions of the gateway the geographic area selection part was a static gif of an Equatorial Cylindrical Equidistant projection on which one could click, then wait for the browser to redraw the image with an "X" on the map, then click a second point, then wait for the browser to redraw the image with a box outlined on it.  Later versions eliminated the wait time by using rubber-band lines. Additionally the map choices were increased to include three different images of the Equatorial Cylindrical Equidistant projection in two different sizes.  Certainly keen, but not much help to the polar community.

In the summer of 1997 the gateway folks included polar stereographic projections in the interface.  For the first time that I know of the polar community had an interface in which they could draw a box around the pole.  Recently the "HTML only" rule has been relaxed and the map interface now also includes a full-on Java "mapplet" which uses an orthographic projection.

JEST: The Java Earth Science Tool is meant to be the V1IMS Java client that will replace the V0IMS X client. For some unknown reason, however, the powers that be have renamed "Version 1" with the much catchier name of  "Version 2".  JEST is still in development and is not publicly accessible.  By the time it is released it it suppose to include polar projections, and those polar projections are supposedly going to be handled properly, but last I knew it didn't.

DIAL: aka "DAAC-in-a-box" the Data and Information Access Link is another Java interface to the same databases accessed by JEST.  As I understand it DIAL was an effort to create a software package that would allow anyone with a web site to put a database search and order tool on their site.   Sounds like a good idea but development seems to have slowed recently.

SEDAC: The Socioeconomic Data and Applications Center is part of the Consortium for International Earth Science Information Network (CIESIN) which is another NASA DAAC and another World Data Center A (WDC-A for Human Interactions in the Environment).  The SEDAC search interface comes in four flavors: SEDAC only or multiple catalog searches in Java or non-Java versions.  The Java versions allow the user to draw a rubber-band box on one of two gif images (topographic or political) of an Equatorial Cylindrical Equidistant projection and use the corresponding range for the search.  The non-Java versions have no map and the user must input the lat/lon extremes of the range information in the text fields.  So the non-Java is still using the Equatorial Cylindrical Equidistant projection - just without the convenience of the drawable interface.

Since CIESIN's focus is on Socioeconomics and Human Interactions, and there is not much of either at the poles, poor polar coverage in the interface is probably not a problem for them.  However, some of the catalogs in the multiple catalog search interfaces do include polar data. Notably the catalog at The National Snow and Ice Data Center.

GCMD: The Global Change Master Directory offers no maps at all in their interface.  The interface is HTML only, and the only geographic area selection capability is four text fields in which the user may enter the lat/lon extremes of a range.  Consequently they too are using only the Equatorial Cylindrical Equidistant projection which, since signs of global change tend to show up at the poles first, is fairly inadequate.

NOAA Server: From what I hear this project came out of nowhere, received huge amounts of funding, did a lot of development work, and then died just as quickly.  There are other rumors, however, that the funding merely got transferred elsewhere for political reasons and the project lives on under a different name.

The interface comes in three flavors: Java, HTML with tables, and HTML without tables.  All three use only the Equatorial Cylindrical Equidistant projection and only support range searches.  In other words they are stereotypical examples of geographic area selection interfaces and hence fairly inadequate for the polar community.

SPIDR: The Space Physics Interactive Data Resource is maintained by the National Geophysical Data Center (NGDC) which is part of NOAA.  SPIDR is not actually a single interface but a collection of data set specific interfaces packaged as a single web site.  Most of the SPIDR interfaces only allow searching the data by date, some only by year.  One exception is the DMSP portion of SPIDR which, in addition to satellite, day, channel, and pass, also allows the user to click on a single point on an Equatorial Cylindrical Equidistant projection and then serves up the eighth orbit browse image(s) closest to that point.

PMEL: The Pacific Marine Environmental Laboratory is also part of NOAA. Early on they developed an HTML only geographic area selection module, and more recently have developed a Java applet. Both support only the Equatorial Cylindrical Equidistant projection and combined they account for about half a dozen interfaces. Among the interfaces using the PMEL applet and/or its HTML only predecessor are: Ferret, EPIC, The Satellite Active Archive, NOAAServer (mentioned above), and the next version of the GCMD interface (mentioned above).

NCDC: The National Climactic Data Center is also part of NOAA. NCDC has their own geographic database search interface for Upper Air Data.  The On-line Access and Service Information System (OASIS) is HTML only, and the only geographic area selection capability is four text fields in which the user may enter the lat/lon extremes of a range.  Consequently they too are using only the Equatorial Cylindrical Equidistant projection.

Oddly the same interface is also available from the Joint Office for Science Support (JOSS) under the name of CODIAC.

NODC: The National Oceanographic Data Center is another NOAA data center that has developed their own geographic database search interface.  The NODC Oceanographic Profile Data Base Online Search comes in two flavors: Java and HTML/JavaScript.   Both support only the Equatorial Cylindrical Equidistant projection.

CEO: The Centre for Earth Observation is a European Communities effort, part of which is run by the Joint Research Centre in Ispra Italy.   While the HTTP based Geo-Temporal Searching (HGS) Tool resides in Europe the mailing list is run out of the Goddard Space Flight Center in Greenbelt Maryland - another NASA DAAC - which is some indication of just how small this community is.

The HGS Search Tool uses the Equatorial Cylindrical Equidistant projection in a Java applet for their geographic area selection.  But HGS is more than just another interface; it is also an attempt to define a protocol (the HGS protocol) which is a bit unsettling given their exclusive use of the Equatorial Cylindrical Equidistant projection.  The protocol itself goes into a fair amount of detail on how geospatial coordinates are to be defined but makes no mention whatsoever of map projections.  For projections, and a large number of other parameters, the HGS protocol defers to the FGDC 1994 Content Standards for Digital Geospatial Metadata which does contain quite a bit of information on projections.

GELOS: The Global Environmental Information Locator Service is part of the G7 Environment and Natural Resources Management project.  Like the HGS interface (mentioned above) the GELOS geospatial interface was developed by CEO.  Unlike the HGS applet the GELOS BBX applet includes polar projections and four different options for defining an area (point, box, polygon, and paint).  What exactly they do with the area once it's defined is still unclear - but I'm looking into it.  Regardless of how the server is set up the BBX applet is still the best geographic area selection applet I've seen to date.

MEL: The Master Environmental Library Query Form was written by Naim Alper of Mar Inc. for the Naval Research Lab at Monterey.  When I first stated looking into the state of the art in web based geographic area selection interfaces, over a year ago, MEL was by far the best publicly accessible interface I saw.  Most of the rest were image maps - and so only gave the user access to specific pre-defined areas - or they were region specific.  A few folks had static gifs in an HTML interface, but at that time MEL was the only decent Java interface.  MEL is also rather straightforward and easy to use.  Moreover, when I wrote to them they wrote back, and were terribly nice people, and gave me access to the source code for MEL.

The only major problem with MEL is that it only supports one projection - the Equatorial Cylindrical Equidistant projection.  But the framework is solid, and they gave me access to the source code, so I spent some time adding polar stereographic and orthographic projections to several custom MEL prototypes.  Those prototypes are accessible from the NSIDC Web Development home page.

Lessons learned from those prototypes are currently being incorporated into the JASPanel (Java Area Selection Panel) that is intended to be a reusable Java module that can do the geographic area selection stuff in any reasonably compatible web based geographic database search interface.  It won't exactly be "plug-n-play" but it'll be better than starting from scratch. A stripped down version of the JASPanel, with fewer options but faster maps, is also online.
 

Those are the ones I know about.  If you know of any geographic database search interfaces on the web that I've missed I'd appreciate hearing from you.  My e-mail address is: swick@chukchi.colorado.edu.