Civil engineer’s perspective
Cote: Honestly, your GIS is probably not fully and truly “model quality” yet, unless somebody was very determined to make it that way from the start. If you want to build a model, you might have a lot of work to do, but on the other hand, there are plenty of nifty software tools to help, so you may be closer than you think. Go ahead and try.

Your GIS is the best possible starting point for modeling. Usually, most of the data you need to build your model is right there. But GIS data is not just used for modeling. The problem with GIS is that because it is so versatile, users can put whatever data, QA/QC, and formatting rules they want into the schema, so no two GISs are truly alike. Therefore, to build the model, you need to adapt to the way the GIS was designed.

To get started, think about the three Cs — Connectivity, Content and Completeness.

Connectivity — By far, the most common problem with GIS is connectivity. For the model to be correct, the pipes must break at every fitting; otherwise, the model sees no connection. For example, in Figure 1, one pipe is selected in red. The pipe must “break” at each intersection.

Content — To build a model, you need certain data such as pipe diameters and elevations, as a minimum. User demands need to be characterized, and might come from GIS, such as meter reading systems, or Traffic Analysis Zones, or estimated from historical use. But the modeler usually has to search out the needed basic data from whatever sources are available.

Completeness — Building the GIS may seem like a never-ending process, but don’t despair. You can get as much as you can from GIS and then manually insert the missing pieces into the model. This gives you an opportunity to see what’s missing from the GIS so you can fix it and later re-import the missing elements into the model.

I prefer a fully detailed every-pipe model, so the GIS and the model should eventually be identical. When you perfect these re-imports, your GIS is truly model quality.

GIS perspective
Jackson: Cote really understands how and when to use GIS, and when to involve GIS professionals in projects. Getting your GIS data ready for modeling should be an intersected event; one that is shared by your modeler and your GIS professionals.

Preparing your GIS data for the model really starts by focusing on what Cote refers to as the three Cs. But not every engineer or modeler is able to “C” and understand why these steps are needed.

Figure 1: The selected pipe (red) must break at the intersection.

So I begin by conducting a thorough GIS data review, or what I call a “behind-the-scenes” look. From there, I compile a facts-and-figures report to share with either my client or supervisor, which explains exactly what needs to be corrected and how much time will be required to make those corrections. This detailed review occurs before the modeling begins, and gives both the GIS and modeling professionals a firsthand view of exactly which GIS tools are needed, and the type of custom scripts necessary to correct the existing data errors. And if needed, the facts and figures in the report can be used as a selling tool to acquire the necessary resources to get the job done properly.

One step that is usually skipped by many who begin this process is understanding how the GIS data was created. Experience has taught us that most utility GIS systems are built using a wide variety of data, such as data transferred from CAD, generated directly from record drawings, heads-up digitized by using a base map such as orthophotography, or field collected using a wide range of GPS technology. Understanding specifically what data came from which methods allows for efficient quality assurance and quality control methods to be incorporated right from the start.

So, if you are about to start this type of project, be sure to begin with knowing what matters. A thorough data review will provide valuable insight on where and how to fix the existing data errors, followed by “C’ing” your way through the connectivity, content, and completeness of GIS data.

Dan Cote, P.E., senior technical specialist with McKim & Creed, specializes in hydraulic modeling, instrumentation design, SCADA systems, communications, and computer programming. He is the developer of McKim & Creed’s comprehensive hydraulic model, Forcemain.
Janet Jackson, GISP, heads McKim & Creed’s GIS activities company-wide. Contact Cote and Jackson at intersect@mckimcreed.com. McKim & Creed is an engineering, surveying, and planning firm.