Information Management for County Infrastructure

Database Design
The foundation of the watershed study is the database that is set up to house all of the information. This database design has to facilitate integration into the county's current GIS database and feed applications such as the stormwater models. To meet this need, the county and Burns & McDonnell decided upon a database developed with SQL Server and Arc/INFOTM with the Spatial Database Engine.

Figure 1 shows the entity relationship diagram for the Mill Creek watershed study. The main linkage for all the information is with the stream network. Since the stream network is the focal point of the database its creation is given a high priority early in the project and will be a valuable piece of information to be added to their information management system.

Data Collection and Survey
The base data used in the study is drawn from the new planimetrics that the county has compiled. These data sources included the roads, hydrology, parcels, buildings, edge of pavement, contours, and water conveyance structures. The new planimetrics have just been completed by the county and provide an up to date baseline of information that is very helpful when doing a watershed study such as this.

The additional data that is required for the study deals with the exact specifications of the water conveyance structures. This information must be surveyed in the field and loaded into the database. A criterion was set by the county that all structures having a contributing area of 160 acres or greater would be modeled. In some cases engineering judgement and field knowledge are drawn upon to extend past the 160 acre criterion to characterize known trouble areas. The location of the structures that meet this criterion is identified with the GIS using ArvViewTM, with the 3-D Analyst and Spatial Analyst extensions. The entire watershed is broken down into six main sub-watersheds to make Stream Network Table data collection easier and facilitate multiple tasks being done at the same time. In some cases engineering judgement and field knowledge were drawn upon to extend past the 160 acre criterion to characterize known trouble areas.

Once all of the locations that need to be surveyed are identified field crews go out and survey each water conveyance structure. At each structure pertinent information is collected dealing with bridge or culvert face dimensions, upstream face cross sections, and road decks. In addition to this survey data, digital photos of each structure are taken and logged with a unique identifier to that structure. The attributes collected at each structure are defined in the database and collected so that they can be loaded into the database with minimal effort.

As the water conveyance structures are surveyed new benchmarks are set throughout the watershed. A total of 153 new benchmarks are set in the Mill Creek watershed. The benchmarks will be beneficial to the county for the stormwater study as well as in future construction and surveying endeavors. The stormwater study provided an opportunity for the county to set these benchmarks when it might not have been feasible to do otherwise. This is one more example of taking advantage of a project to supplement an information management system.

Hydrologic Data Development
As the field survey work is being completed the data development to support the modeling is going on in the office. The hydrologic modeling for this study is done using the U.S. Army Corps of Engineers Hydrologic Modeling Center model HEC1. This model is used to determine the flows in the stream network throughout the watershed. In order to feed this model Arc/INFOTM and ArvViewTM are used to manipulate the county's planimetric and elevation data.

The first step in generating the data for the HEC1 model is to take the hydrology from the planimetric data and create a fully connected stream network. This was started during the process to determine which water conveyance structures would need to be surveyed. At this point, though, the stream network extents that were determined from the initial 160 acre criterion are pulled out and cleaned. The cleaning of the stream network involves heads up digitizing a stream centerline in Arc/INFOTM for every portion of the network that had a width greater than 10 feet. This is necessary because the planimetric hydrology file contains water edges for every stream with a width of 10 foot or greater and not a stream centerline. Coupled with this fact was that whenever the stream went through a conveyance structure or underground it was not captured in the hydrology file. The result of all this work is a stream network that is fully connected and topologically correct for the entire watershed. The stream network will be delivered to the county as an Arc/INFOTM coverage so that they will not have to go through this exercise of heads up digitizing the next time they do any type of stormwater project for this watershed.

With the stream network complete, focus is turned towards the analysis of the watershed terrain. As mentioned earlier, ArvViewTM Spatial Analyst and 3-D analyst were utilized to help determine what water conveyance structures should be surveyed. In order to do this a triangulated irregular network (TIN), flow direction grid and flow accumulation grid is developed for each of the 6 sub-watersheds in the Mill Creek watershed. The development of these data layers is done with the Hydro tools extension for Spatial Analyst. In addition to the grid files a point file is created for each sub-watershed that contained the location of the desired pour points. The pour points are set at water conveyance structures and major stream confluences. Additional pour points may be set other places to split up a large sub-basin or better delineate a particular area. The pour points and grids are used with an avenue script to delineate the sub-basin area contributing to each pour point.

Once all the sub-basins are delineated the integral data that will be used by the HEC1 model is extracted. The bulk of the remaining data generation deals with the determination of the runoff curve number for each of the sub-basins. Some of the other data that is developed includes: maximum flow path lengths and slopes, closest rain gauge station to each sub-basin, and stream centerline lengths and slopes.

Two options present themselves for determining existing conditions run off curve numbers. One option is to use a landuse coverage to determine the percent impervious areas or alternatively, use the planimetric data to determine a percent impervious. The latter of the two options is chosen because using the planimetric data will provide a more accurate account of the actual impervious areas in the watershed. Therefore, the coverages containing roads, buildings, driveways, and edge of pavement are combined to yield a true impervious area coverage for the entire watershed. With the impervious area coverage built it is then combined with the soils data to determine an existing conditions run off curve number for each sub-basin.

The planimetric data is also used to develop a set of representative percent impervious tables for the Mill Creek watershed. The tables give percent impervious values for each landuse in the study area and can then be used in determining the future conditions run off curve numbers. This is necessary since only landuse will be available to determine future conditions information. This method will calibrate the percent impervious values to the historical development trends in this watershed and will be better than using a set of default book values. The tables will be provided to the county as a deliverable so that they will be able to reproduce everything that was done in this study as well as provide the information to anyone who does future studies in this watershed.