The Biological Resources Research Center is developing a web-accessible Biodiversity Atlas for Nevada and the Great Basin. We want to present as rich a cartographic experience as possible for the wildest number of users. This necessarily precludes the use of an out-of-box ArcIMS solution, since not all browsers/platforms (particularly Macintosh browsers) will reliably display pages generated by the ArcIMS. We are developing a solution using ASP and JavaScript for ArcIMS that is accessible to all standard Wintel and Macintosh browsers providing much better performance and wider accessibility to the standard GIS data formats through SDE. We will discuss methods, problems and solutions.
The Biological Resources Research
Center’s mission is to conduct scientific research and planning efforts to
preserve the distinct biotic diversity of Nevada while providing for economic
viability and other needs of its citizens.
Most of the scientists on staff are ecologists and/or conservation
biologists, both of which are largely field disciplines and therefore have
geographic components. The BRRC’s GIS lab assists these scientists to better
understand the geography of their respective study areas, and serves as a
repository for most of the geographic data collected in these efforts.
One specific assignment for the
BRRC’s GIS lab has been the maintenance of the Nevada Mountain Atlas. The
Nevada Mountain Atlas is a series of static web pages that contains species
lists and summary geographical statistical information, like high and low
elevations, relief, area, and highest peak, about the 314 named mountain ranges
(McLane, 1978) in the state of Nevada. The GIS lab soon began getting webmaster
emails detailing what various deficiencies were in the quality of information
available on the site. Some species’ were listed using outdated taxonomy or
completely erroneous taxonomy; picture links would show images of the wrong
species; low elevation species would be listed for high elevation sites; etc,
etc. We did our best to correct these problems as we became aware of them, but
the task of keeping 314 static pages full of fairly repetitive summary
information both up-to-date and error-free is daunting at best. We decided that
we needed to present the data dynamically as it appears in our databases,
rather than having 314 static web pages that need periodic review,
recompilation and constant revision due to errors. Additionally, we felt that
the Mountain Atlas concept was flawed in the context of the BRRC’s mission.
Between most of the mountain ranges lie the cold steppe and desert valleys that
most people associate with Nevada, and contain many of the species that the
BRRC wishes to research and/or monitor. The BRRC is tasked with conducting
ecological research over the entire state of Nevada, not just its mountains.
The new concept is called The Nevada Biodiversity Atlas. Ultimately, ArcIMS was
chosen for displaying the Atlas.
There are several categories of
data that we wish display: base layers like shaded relief grids, state and
county boundaries, map quadrangle boundaries, PLSS boundaries, roads, lakes,
streams and rivers; species accounts (points that represent places where
type specimens were collected and subsequently curated); and land cover
coverages and grids, like DEMs, GAP vegetation (vegetation classified into
communities), geology, soils, and the basins and ranges that provide
context for all of the other data sets
(this instead of the of the old model that was centered on the 314 named
mountain ranges). With IMS, the user can make some simple spatial analysis’
with these data, like what species’ occur in a particular valley or set of
valleys, or what counties does a particular river cross? This is useful to both
the scientists of the BRRC who can make simple analysis’ for themselves
immediately, and for members of the general public who may wish to simply know
what maps they will need for their trip to the Ruby Mountains or who may wish
to know where they might find pallid bats. Later implementations will include
data on various research projects in which the BRRC is involved.
We have several design goals.
Create an online mapping application to display the Nevada Biodiversity Atlas.
Make the Atlas as platform independent as possible. Integrate the look of this
application with the rest of the BRRC website (which itself is undergoing a
major design upgrade). Integrate the Atlas with other areas of the BRRC
website: we want to be able to give geographic context to all areas of the
site. Enhance the functionality of tools provided by IMS to make them fit our
own conception of how such tools should work.
It is not difficult to create an
IMS application using ArcIMS Designer, but the application that it creates makes
extensive use of JavaScript. All requests made to the ArcIMS server are made in
XML, which is built on the client side using JavaScript. ArcIMS Designer is a
good tool for those who need an application to be created quickly and easily
but it has some disadvantages. For instance, for every operation an XML request
string must be generated on the client using JavaScript. This is slower than
using methods from ActiveX connector object model as we just need to call
the appropriate method instead of building a request string. This also prevents
the application from working on various Internet browsers for the Macintosh; we
have many Macintosh users.
To make the application run faster
and make it more platform independent, we used Server Side scripting which
generates HTML and sends it to the client. In ASP we can use both server and
client side scripting. Most of the HTML is created on the server and
validations of user input are done on the client. This helps make the
application run faster. The application created using ASP will run on most of
the various Explorer and Netscape browsers on both PC and the Macintosh.
Using the ArcIMS ActiveX connector
Object Model and programming it using ASP was the best available solution for
us with which to build the application. Using the ActiveX object model we can
easily customize the tools according to our needs and build new tools making
use of different methods available in the object model.
There are other advantages of using
ASP. ASP takes less time to write and execute. It is multithreaded, allowing a number
of concurrent users. ASP can also be used to maintain user sessions. Using this
we can uniquely identify each user session and depending on type of the user we
can restrict or provide access to the sensitive data in our application. We can
set session timeouts with which we can terminate a session if the user is idle
for a given period, thus reducing unnecessary load on the server.
The next decision was determining where
to store the data required by the ArcIMS server. The data can be stored as
shape files or in a geodatabase using ArcSDE. There are advantages of using
ArcSDE for storing the data. With ArcSDE we can work directly
with spatial data managed in a RDBMS. Performance is enhanced through storage methods that
provide a fast and compact representation for spatial data. We can move data
between different databases using ArcSDE’s data import-export capabilities. It
defines a single logical model for spatial data implemented on top of the
particular physical database. It also manages the integrity of point, line and
polygon information added to the database and does not allow ill-formed feature
geometry to be inserted. These factors, plus SQL Server being both relatively
inexpensive on an academic license and somewhat less intimidating to novice RDBMS
managers than other “real” RDBMSs, led us to the decision to use ArcSDE with
SQL Server.
The layers were divided into two
groups one having layers with sensitive data and the other having layers with
non-sensitive data. In identify and query operations the
sensitive layers were filtered so that only non-sensitive data would be
displayed to the user protecting the sensitive data. In this application, buffer
operations return a list of species that occur in a particular polygon or set
of polygons. Since specific locations aren’t given, the sensitive species
layers are not used to filter sensitive species.
Some of the interface functions from the ArcIMS HTML template have been changed to suit our needs.
1. PanningPanning can be done in eight directions, instead of the standard four on the vertical and horizontal axis’. However, the hand-pan tool is still probably most useful for most users. 2. Zoom-InZoom-in is limited to a maximum scale such that the user cannot zoom in so close that terrain details become obscured by pixelization. 3. Zoom-OutZoom-out is limited to the map extent as defined by the current window size. 4. IdentifyOur identify tool has been enhanced so that the user can draw a rectangle around multiple features and get information about all of them as well as being able to click on single feature. 5. BufferThis is an important tool that can be used to find all the features in a particular polygon layer (basin or range). Initially a buffer is created on a basin or a range. Then we select layer and a list of all the features of this layer present in that basin or range is displayed. 6. Window sizeWhen the IMS browser window is re-sized, the map image retains its focus and the extent remains nearly the same (changing the window size usually means that the aspect ratio is also changed, so the extent changes to accommodate that.). We have been unable to find another IMS implementation that does not return to the full map extent when the window size is changed. |
Buffering:
The goal for the buffering
tool is to be able to select a feature or group of features and have IMS be
able to both display the locations of selected
features within the buffer
parameters AND be able to deliver a neatly formatted report containing the
summary results of the buffer operation. As this is being written, the report
function works, but the buffer operation is very slow compared to the buffer
tool provided with the standard HTML IMS viewer. Currently, we are only
allowing buffers on one feature at a time. We anticipate being able to solve
this problem, and hope we have a solution in time for the formal presentation
of this paper at the 2002 Esri User Conference.
The first version of the application is complete and is online available at http://134.197.55.92:90. It is able to display the data sets that make up the Nevada Biodiversity Atlas. Since the application was developed with ASP it is a lot faster in loading and navigation as compared with other similar applications built with the ArcIMS Designer. We have provided some enhanced tools to the interface. All spatial data is stored in SQL Server 2000 using ArcSDE. The application has been tested (and continues to be tested) and works well enough that we provide links to it from the BRRC website. The application displays in standard PC and Macintosh browsers. The application displayed the web pages correctly at any screen resolution and/or browser window size. The map image retains its focus and stays as close to the scale at which it was previously displayed as possible given the probable change in aspect ratio when the browser window is resized.
The application was faster as compared with others and it ran on most of the browsers on PC as well as Macintosh. The application supported all screen resolutions and resizing browser window preserves the extent of the map.
Esri website www.Esri.com
McLane, Alvin R. 1978. Silent Cordilleras. Reno, Nevada. Camp Nevada
Alex Homer, Dave Sussman, Brian Francis. 2000. Professional Active Server Pages 3.0, Wrox Press Ltd.
Using ArcIMS 4, Esri. Redlands CA Esri Press
Customizing ArcIMS ActiveX Connector, Esri. Redlands CA Esri Press