In Chester County, Pennsylvania, the Chester County Water Resources Authority (CCWRA), with support from Camp Dresser & McKee and Gaadt Perspectives, is developing a comprehensive water resources management plan. The plan, called Watersheds, has many facets. It includes analyses of groundwater and surface water quality and quantity, water supply and wastewater disposal issues, resource protection and its relationship to land use, pollutant loading analyses, future population impacts, historic and recreational aspects of water, among other issues. The approximately 1400 square mile Study Area encompasses 78 subbasins within 21 watersheds that cut through nearly 160 municipalities and various geologic formations and are located in six Pennsylvania counties and parts of Maryland and Delaware. Thus, the huge amount of spatial data needed for the Plan created a challenge, and well-planned data management using ArcView^®, ArcInfo^®, and Microsoft Access^® was critical to our success.

Spatial data analyses included intersecting several data coverages, such as municipal boundaries, subbasin boundaries, population census blockgroups and land use. There were several major data evaluation elements including the development of current and future water balances/budgets, current and future pollutant loadings, present and future land use analyses, relationship between the various water needs and definition of problems/concerns across the Study Area. There also were several major issues that needed to be evaluated and integrated in this regional watershed project. The major issues included:

• Hydrologic Cycles
• Water Use
• Land Use
• Wastewater Service
• Demographics
• Storm water Systems
• Water Supply Systems

In the Study Area, water flows into two major river basins, the Delaware River and the Susquehanna River basins, or directly to the Chesapeake Bay. Watersheds, as we have defined them in this study, are the medium sized drainage areas labeled with their primary creek names, such as White Clay and Brandywine Creeks in Figure 1. The smallest drainage divisions in this study are the 78 subbasins, which range from an area of roughly 4 to 50 square miles. The whole Study Area covers approximately 1,400 square miles.

Counties and municipalities are key stakeholders in this regional plan, therefore, their location and boundaries were also an important data element. There are approximately 160 municipalities in the Study Area that are shown in Figure 2. The complexity of watersheds and subbasins crisscrossing municipal, county and state boundaries required the use of a geographic information system (GIS) for the plan. For this we chose ArcView^® and ArcInfo^®.

Regional watershed planning primarily involves managing and evaluating two types of data sets, spatial data and tabular or characteristic data. Much of the spatial data for the plan came from GIS map coverages. These maps included polygon data sets such as public drinking water and sewer service territories, geology, land use, zoning, population census blocks, soils, and municipal boundaries. Point source data maps included water withdrawal, recharge and discharge point locations, water quality measuring points for numerous parameters (metals, total suspended solids, phosphorus, pesticides, volatile organic compounds, nitrates/nitrites, and others), stream and rain gages, observation wells, areas that flood and many more. Line data included streams, roads and pipelines. The other basic element was tabular data consisting of numerous tables of information. Examples of these data include water use amounts for all the discharge, recharge and withdrawal points, municipal population figures for 1998 and future years, water quality data, precipitation data, groundwater flow data and much more. Figures 1 and 2 show spatial data and Figure 3 gives an example point data set.

We gathered the data from various federal, state, and local government agencies, existing water studies in portions of the Study Area, state water use reports, internet web-sites, universities, meetings with stakeholders, and surveys/questionnaires sent to municipalities, water service providers, sewer service providers, the public, watershed associations and counties. The data was provided in a wide array of formats including Microsoft Excel^® tables, text files, ArcInfo^® and ArcView^® maps (in several projections and different datum, some with complete documentation and tables, others with little data), paper maps, written reports, archaic government computer formats, written survey responses, and verbal discussions. Each data set had different levels of quality, completeness, and usefulness. All these factors made up front planning and integrating Access^®, ArcInfo^® and ArcView^® critical.

AAA Integration

Our primary GIS "workhorse" was ArcView^®, an excellent program for viewing and mapping data. We used ArcInfo^® when we needed more complex spatial analyses such as intersects. Based on our past experience we used a proven and robust database program Microsoft Access 97^® to compile, query and report all tabular data. We imported and exported data between ArcView^® and Access^® as needed that enabled efficient visualization of compiled and summarized database information.

Regional watershed planning typically requires evaluating current and future characteristics of the watersheds, developing a problem list and arriving at suitable alternatives to solve the problems. Our approach was to develop one database with all the data required for characterizing both current and future states of the watersheds. Our first step was to address the data management needs of all the tasks and carefully build a good set of relational tables and GIS maps. Although the initial task may appear to not provide any results for a while, the benefits of such an approach become obvious once the evaluation and latter tasks begin. The "what if" capabilities of integrating ArcView^® and Access^® are enormous and enabled us to rapidly respond to the numerous data evaluation requests from the watershed planning team.

Once we began gathering the enormous amounts of data for this project we processed it into uniform format and double-checked the quality of all the data. For the spatial data this meant bringing all coverages into the same datum and projection (state plane, Pennsylvania south with 1983 datum), checking the data in the tables against other data sources and plotting spatial data we received. Information we plotted in ArcView^® included withdrawal points, surface water discharges, recharge points and others that we received with latitude and longitude data.

Once we had high quality base coverages we got into the heart of the spatial analyses. We had several base maps, such as subbasin boundaries, municipal areas, census population blockgroups, water and sewer service territories, land use, and geology. We used ArcInfo^® to do double, triple and quadruple intersects with these coverages, such as subbasins, municipalities, census blockgroups and water service areas. We then would export the table to Access^® where we would use the data in various ways, such as to estimate the number of people by municipality who live in areas not served public water.

Two additional components of the Watersheds project were subbasin level water balances and preliminary pollutant loading estimates. Typically, water balances or budgets are very cumbersome and require significant amounts of time and effort. All of the precipitation data, water withdrawal and recharge data and run off characteristics need to be compiled on a subbasin by subbasin level and evaluated. Often this can require one to two months of work. In our integrated approach, compiling all the data into one database enabled us to set up queries that automated the task and we were able to evaluate and develop the water balances in a matter of two to three weeks for all 78 subbasins. We then would export the water balance information back to ArcView^® to spatially display the data, for example the degree to which the subbasins were over-drawing the water resources.

We also used a simple Microsoft Excel^® based model to estimate runoff volumes and pollutant loading from runoff, baseflow and septic systems. The base unit for the model was the intersect of the subbasins with the municipal boundaries. The model required processing landuse data, soil characteristics, point sources, septic systems and precipitation data. When Excel^® finished running the model we brought the model output back into the Access^® database to process it, then exported the processed data to ArcView^® for mapping. We were able to quickly develop summary tables and ArcView^® maps to display pollutant loading and runoff amounts. Figures 4 and 5 are examples of these maps and show how total suspended solids and runoff vary over the Study Area. We were able to visually highlight the watersheds of concern and also evaluate the future watershed scenarios. Again, the combination of ArcView^® and Access^® was the key to our success.

The End Result

The combination of ArcView^®, ArcInfo^® and Access^® gave us an excellent spatial and tabular data set that has enabled us to fairly easily develop a complex report. The powerful combination has reduced the level of effort associated with the data compilation, provided more time and resources for data evaluation, and helped us to develop a sound watershed planning guidance for the entire Study Area.

As an example, one section of the Watersheds report is designed to provide a framework for multi-municipal watershed based plans or "Integrated Water Resource Plans"(IWRPs). To make the strategies as useful as possible to the municipalities charged with implementation, we provided two levels of recommendations: general recommendations that were applicable to most of the study area, and more watershed specific recommendations for each of the 21 watersheds. In these sections we evaluated and summarized many things such as: areas that need public water or sewer based upon future growth plans, sensitive areas, existing pipelines, and current needs; stream segments that need to be better protected because of current land uses and population impacts; and recommended future water sources for each watershed emphasizing water use within the same watershed. Using the "triple A" combination made these recommendations possible in a relatively short amount of time.

Positives, Pitfalls and Points to Remember

With all our careful up-front planning and powerful programs we still learned some lessons, encountered some problems, and learned some tricks for even smoother run projects in the future. Without ArcView^®, Access^® and ArcInfo^® we could not have completed a project of this scope as easily or as quickly, if at all.

One problem we encountered early in the project mystified us initially. Figure 6 shows the subbasin map of the study area, and Figure 7 shows the general geology of the area. Figure 8 shows their intersect using ArcView^® 3.2 with subbasins as the overlay, and Figure 9 shows the same intersect with geology as the overlay. As figures 8 and 9 show data was lost when using ArcView^® to do the intersect. Once we switched to ArcInfo^® for all intersects this problem went away, and we discovered that ArcInfo^® is the tool of choice for complicated intersects. Figure 10 shows the geology/subbasin intersect using ArcInfo^®.

Another lesson we learned is never to change a primary piece of spatial data late in the game if at all possible. Several months into our project we made a minor change in the subbasin map, combining some small subbasins and clipping a subbasin from the edge of the study area. This created a full week of extra work since we had to intersect all the maps again using ArcInfo^®, re-import all the related data and do all the queries and related analyses again.

A habit we developed during the project was to carefully document where the data came from, from whom it came, what we did with each piece of data and what data we used for each calculation. We did this to avoid redundant work as the project progressed and to help in our final project data documenting. It is easy to forget several months later, for example, if you already imported and compiled certain pieces of data, or which coverages were used in an intersect.

We also found it useful to periodically make copies of the databases and keep older versions at different points in time. Several times we went back to get old data we either deleted or changed so much we needed the old version of the table or map again.

Finally, automating the data management tasks where possible so we could take the "cookie cutter" approach saved us weeks of work. Keeping table and field names consistent in Access^® helped speed up processing time. For example, to process ten different water quality parameters we used consistent table and field names, this allowed us to copy and paste existing queries and reports and change only their references in the database. Similarly, keeping a core of base maps of high quality for all visuals made it easy to quickly create new figures or maps as needed.

Conclusion

Regional watershed planning projects involving several watersheds and subbasins have a multitude of tasks, each with demanding and comprehensive data management needs. Using an integrated data management technique and combining all the data management activities has proved to be crucial and vital to the success of such large-scale efforts. Careful up front planning and the building of a well thought out, relational data management system can prove to be an enormous benefit. Not only will such an approach reduce the time and effort required to execute the planning tasks, but also enhances the accuracy, reliability and quality of the data evaluation process. The combination of ArcView^®, ArcInfo^®, and Access^® provides an excellent, high-speed tool for developing water balances, for watershed pollutant loading models and for watershed current and future evaluations.