The Wisconsin Department of Resources (WDNR) is currently working on a 24,000 scale (24K), digital, spatial hydrography layer for Wisconsin. As part of this work, the Surface Water Integration Committee (SWIC) contracted with GeoAnalytics of Madison, WI, for the construction of a prototype data editing, query and analysis application. The intent of the project was to test the dynamic and geocoding characteristics of the proposed final hydro design.

Many of the 24K hydrographic structures being built by WDNR are designed for the purpose of determining interrelationships between various environmental data associated with surface water features. This effort is part of a much broader initiative called the Surface Water Integration System (SWIS). Water is the unifying element in this model and water flow is the method through which one event can influence or initiate other events.

The SWIS prototype application for Arcview is designed to fulfill a number of criteria based on the needs identified by the potential SWIS user community. First and foremost, the prototype provides users with tools to query and analyze hydro related data using spatial methods. The foundation for this functionality is the "Phase 2" 24k Hydro spatial data set. Phase 2 incorporates advanced linear and areal features that provide an intelligent hydrographic framework. Data "events" can be attached to this Geographic Information System (GIS) and related to one another spatially.

Another major criteria of the SWIS prototype for Arcview, is that it be based on the DNRview data model and application. DNRview is a data distribution and implementation effort being developed by GEO Services Section. It provides a standard mechanism by which DNR staff can easily access resource information throughout the state. The SWIS prototype demonstrates ways in which applications can be built upon the DNRview foundation.

To accomplish this integration of data and applications, certain low-level enhancements were made to the DNRview model. DNRview is based on an underlying structure of standard directories referred to by system variables. This allows for a maximum consistency and flexibility that is required by the dynamic nature of the current computer environment. The model was altered for SWIS so that developers could add their own sets of application variables. A means was also provided to access extensions outside the standard Arcview installed directories. These changes have now been incorporated in the production version of DNRview.
 

DNRview Add Ons

The SWIS prototype has been designed as a series of independent modules, each providing a subset of the total functionality. These modules are built as Arcview Extensions and can be used in combination or separately depending on the user's needs. Extensions allow developers to make incremental upgrades by only replacing the module file that has changed. The enhancements are automatically incorporated into existing Arcview projects that use those modules.

One common way developers customize Arcview is to add menus, buttons and tools to the standard interface, giving users access to specialized programs. DNRview has a number of these additions designed to enhance staff productivity and efficiency. In order for DNRview to serve as a basis for further customization, the available screen real estate for these controls needs to be economically managed. The SWIS prototype has implemented the Dialog Designer capabilities of Arcview as a way to provide each additional add-on with its own control space.

The Dialog Designer allows the designer to create window dialogs outside the Arcview window. These can be moved anywhere on the monitor display and opened and closed as needed. The SWIS prototype has a standard set of control windows for each application. These dialog windows provide access to custom menus, buttons and tools without infringing on the default Arcview control space.

All SWIS extension functionality is initially accessed from a single menu added to each of the default Arcview GUI menubars called "Apps Menus". Each application is listed as a choice within this menu. Once the root menu is loaded, the user can select any of this dialog's menus, buttons or tools as a means of accessing additional dialog menus or custom functionality.
 

SWIS Prototype Modules

There are four extensions or modules in the SWIS prototype. The first of these modules is "SWISview". SWISview is a companion to DNRview, offering access to spatial data sets that are not part of the typical distribution, but that reside within a DNRview-like directory structure. These data sets include the WISCland land use layer, Endangered Resource layers, and preliminary 24k Hydro spatial data.

The second module is "Image Tools". With this extension the user can organize and manipulate image data sets used in conjunction with SWIS operations. Image Tools contains additional functionality which lets the user select and apply pre-configured symbol legends to image themes without the use of Spatial Analyst. The user can easily create image catalogs and combine them into spatially indexed libraries so that images stored on different removable disks can be easily located and loaded.

Module number three is "Point Locator", providing a set of tools for loading, creating, and editing the location of points stored as coordinates in tabular records. Much of the WDNR environmental data is stored in databases and includes the coordinates of the collection points. Arcview can take XY fields and convert them to shape points. The SWIS extension contains methods for adjusting the location of these points, updating coordinate information in the database, and converting the points into hydro-linked events.
 

Upstream, Downstream

MapQry, the last module for SWIS, builds on the advanced spatial features proposed for Phase 2 of the 24k Hydro data model. This data structure lets users associate point and linear data sets as events along the hydro routes and and as coincident with hydro regions. MapQry also incorporates the data conventions of DNRview and SWISview. This means that these two modules do not need to be loaded to access any of their layers from within MapQry

Figure 1

The MapQry module gives the users a way to ask questions and compare data in new ways, based on the ability of routes to "know" the spatial relationships of events located on them. Thus, users are able to ask what hydrographic elements are upstream, how far, and how much. Or they might want to know the range of activity within a watershed, or what streams and lakes are downstream of some point. These capabilities have been implemented in preliminary form by the SWIS prototype so that users can evaluate the tools and the data model.
 

Data Components

At the core of Map Query lies the 24k Hydro Phase2 data. For the purposes of the prototype, two watersheds were prepared to meet the current design specifications for Phase2. They were structured into layered datasets of increasing abstraction and complexity, beginning with the basic arcs and polygons, and progressing up through dynamic segmentation routes and open water regions to linear water body events and the network traceable routes.

This structure offers a many faceted geographic entity that can be viewed in many different ways and can be used in a variety of interactions with other geographic data. Much of what the user wants to know about the hydrography is how it affects or is affected by environmental data that has been collected in and around the water. By processing the "geo codes" stored with this data, SWIS creates a spatial link to all of the rich environmental data stored in the department databases.
 

Data Consolidation

The Point Locator gives users the ability to create numerous point datasets for use in conjunction with 24K hydrography. Map Query uses a device called a "rule base" to consolidate and standardize access to these layers. Once the data is described within the rule base, the application program can find and manipulate it in standard ways.

The Theme Configuration dialog offers users the ability to register themes in the active view with the rulebase. At the same time the user can identify any tabular items within each dataset that are most likely to be queried. This simplifies query construction by limiting the choices to only those values that are important. Map Query is delivered with all of the layers available in DNRview and SWISview already configured in the rule base. The SWIS "Add Theme" dialog provides a scrolling list of all these theme names in a one-stop-shopping Arcview theme loader.

User data, in comparison to the corporate datasets, is dynamic and created on the fly. By opening the User Data Configuration Dialog, these adhoc layers can be loaded into the rule base where they can be accessed in the same way as the corporate layers. An additional feature of user data configuration is the ability to associate the geographic features with items in a related table based on some primary/foreign key relationship. These related items can then be queried along with the base feature items in the Map Query dialogs.
 

Query/Analysis Options

There are three major ways that data can be queried or analyzed in the MapQry module. First, the items that were identified as queriable are available to the "Search Theme" dialog. Any of the rulebase layers are also available for making spatial reselects of any other defined layer through the "Select by Theme" dialog. The third way to query is to perform a network trace and then use it in a spatial reselect.

The Network Query dialog provides the means to create a transport fragment from 24k Hydro routes and then apply it in a spatial query. The fragment is defined as either an upstream trace from a specified point, a downstream trace from a certain point, or a trace between two specified points. The length of the upstream and downstream traces can be limited to a particular distance in miles or to the end of the reaches.

The network trace algorithm was developed independently of the Network Analyst available from Esri. The WDNR wanted to evaluate how well specific network functionality could be provided through Avenue programming. The SWIS trace tool produces he desired segments, but at a much slower response time compared to compiled programs. The Environmental Protection Agency had a custom trace tool built in C, which is then called by Arcview. This program produces a result in a more reasonable time frame.

Once a trace is created, it can be used to select features in other themes that are in spatial proximity to it. The trace, however, is a complex, linear shape entity referred to as a linear event. As such, it is linearly aware of other events associated with the hydrography, in particular, the environmental point events created by SWIS users. Events are really just tabular relationships that are interpreted by Arcview as spatially displayable locations. Based on this, one can imagine a linear query taking place solely within the corporate relational database and returning the results to Arcview for display.

Taking this a step further, if all environmental locations were stored as hydro event tables in the corporate database, a linear query based on the relationship of point events to a linear trace event could return all environmental data associated with that length of stream. None of the separate point themes would need to be loaded in Arcview, and only those points that were returned by the query would need to be managed and displayed by the view.

As a test of this scenario, the SWIS user test events were loaded into an Access database along with a trace event table. Using the tools provided by the standard Arcview extension, Dbaccess 2.0, an Arcview table was constructed based on a SQL join statement executed against the Access database (Figure 2). The table contained the point events that met the criteria of being associated with the same stream segments as the trace (Figure 3). As point events, these records could be displayed spatially and classified by type (Figure 4). By means of their unique ID's, they could also be joined back to the wealth of environmental data stored in the corporate database.
 

Figure 2

Figure 3

Figure 4

Conclusions

The WDNR has used the SWIS prototype to evaluate ongoing 24K hydrography layer development as well as to guide plans for future user interfaces. As time goes by and newer technology becomes available, these plans may be substantially modified. One thing that will not change is the goal of finally integrating he spatial view of environmental elements with their tabular view. This will be possible from within the entire enterprise IT framework and not just on isolated GIS workstations.
 

Pete Thum
GeoAnalytics, Inc
Madison, WI 53704
Ph 608-241-7100, Fx 608-241-7116, coryj@geoanalytics.com

Jim Cory
GeoAnalytics, Inc
Madison, WI 53704
Ph 608-241-7100, Fx 608-241-7116, coryj@geoanalytics.com