26 • May/June 2004 • Southwest Hydrology

Peter Martinez, Lance Kuhler, P.E., and Andrew Messer, R.G. – URS Corporation, and Scott Estergard – U.S. Army Corps of Engineers

The U.S. Army Corps of Engineers, the city of Phoenix, and URS Corporation are conducting a

feasibility study for a stretch of the Salt River referred to as Rio Salado Oeste (Salt River West). The study aims to address issues of native riparian ecosystem restoration, flood damage reduction, and passive recreation. The project area encompasses approximately eight square miles and lies between two other concurrent environmental restoration projects, Rio Salado and Tres Rios.

The GIS was a critical component in the environmental restoration process. Once all data layers were gathered, they were projected to the project coordinate system. More than 50 datasets were incorporated, including: elevation contours, aerial imagery, political boundaries, land ownership, existing land use, soils, storm drains, discharge points, flood zones,

treatment plants, and general vegetation. GPS technology was used to verify discharge points and delineate individual vegetation stands.

When the necessary datasets were in the correct coordinate system, their database attributes were evaluated and used to develop a ranking criterion for the analysis process. Some datasets such as vegetation were buffered in order to portray more realistic geographic phenomena. Others were merged to produce a continuous dataset with varying values over the project area, such as the 5-, 10-, 50-, 100-, and 500-year flood zones. The project team, which included biologists, geologists, and hydrologists, developed weighted values for each of the datasets used in the analysis. These weighted values determined the importance of each dataset used during the analysis process.

ESRI’s GRID was used to generate groundwater elevation data, while ArcGIS was used to generate map products. ESRI’s Arc Macro Language (AML) and ArcInfo were used to write and process the analysis routine. The AML routine

performed spatial overlay analysis of the different datasets to generate an output dataset with the sum of the weighted values of the individual polygons as they were compiled. At the end of the routine, all polygons with the same values were merged. The final analysis dataset was then used to determine suitability for implementing restoration measures throughout the study area.

Using GIS provided several benefits to the analysis. It assured that: 1) the process and methodology were uniform, objective, and systematic; 2) the data were edited efficiently; and 3) the analysis could be rerun relatively quickly to evaluate different scenarios. GIS also proved ideal for generating presentation boards, maps for reports, and area summary tables based on the results. As the project enters its second year, the GIS database will be updated and expanded as needed. The GIS is also fulfilling an early goal of providing a storehouse of project data to be shared with other agencies and researchers over the life of the project and beyond.

Contact Lance Kuhler at Lance_Kuhler@urscorp.com.

GIS as a Tool for Riparian Habitat Restoration

Riparian habitat delineation in theRio Salado Oeste area using GIS.

May/June 2004 • Southwest Hydrology • 27

Regional-scale groundwater models typically cover multiple basins and span several decades

of simulation. For regions with complex geology and natural and man-made hydrological boundaries, a significant volume and diverse types of data must be compiled, explored, and visualized before a conceptual model can be constructed. During model implementation, data from different sources must be translated into the model-specific format. During model calibration, data quality assurance/quality control is needed to identify hidden errors associated with the input data and model conceptualization. Each of these steps requires the constant display and re-assembly of datasets to determine the accurate position and value of significant parameters, to reveal cause/effect relations of critical components, and ultimately to match simulation results with independent measurements. For distributed-parameter models with multiple simulation stress periods and a variety of comparisons, this process is iterative, time-consuming, and technically challenging. Powerful data processing

and visualization technologies can facilitate efficient development of regional groundwater models, and were used to build a model of the Twentynine Palms area of California.

Twentynine Palms is in the Mojave Desert, about 130 miles east of Los Angeles. Multiple studies spanning several disciplines have been completed in parts of this area. Water managers wanted a regional groundwater flow model to evaluate long-term water availability and optimize water operation and management of several basins. To develop such a model cost-effectively, an efficient data-processing tool or platform

was needed to manipulate the large volume of data and models that use a variety of types, formats, and projections.

More than 10,000 groundwater-level and water-quality measurements collected over a 50-year period were retrieved from the U.S. Geological Survey’s NWIS database. Combined use of GIS and visualization tools (such as Tecplot) allowed the data to be swiftly organized, interpreted, and depicted using water-level and water-quality maps with a variety of themes for different uses. Overlaying and cross-correlating water-level maps with coverages of mapped faults, well fields, and surface-water features allowed rapid identification of the major spatial features controlling the natural variation of hydraulic head, such as faults and lakes, and the temporal changes induced by human activities such as pumping. Three local groundwater models that originally had different units, grids, and orientations were transformed, unified, and visualized into the same projection as the regional model (see map). By comparing and integrating information from these models, the regional groundwater flow budget and flow direction were quickly derived.

Data from other disciplines also were visualized or translated into the regional groundwater model grid using GIS. A mean recharge matrix was created from an infiltration model, and the bottom elevations of the regional model were defined on the basis of gravity data from a geophysical model. Additionally, GIS was used to overlay an aeromagnetic map on a geologic map to refine locations of possible fault-defined flow barriers, and to overlay the geological map with coverages of arsenic and chromium concentrations to correlate degraded water quality with certain geological units.

Development of the regional groundwater flow system model was significantly expedited using GIS and other visualization tools. Results from the preliminary model proved to be reasonably accurate in matching observed groundwater level distribution.

Contact Zhen Li at zhenli@usgs.gov or Peter Martin at pmmartin@usgs.gov.

GIS for Groundwater Modeling and VisualizationZhen Li and Peter Martin – U.S. Geological Survey

Regional groundwater flow budget and flow direction were derived by integrating three previously developed local models.

28 • May/June 2004 • Southwest Hydrology

One might think from other examples in this issue that GIS is useful only for state-wide analyses or for

constructing regional models. Not so! GIS is also used to great advantage on small and low-tech projects, or solely for the production of high-quality presentation graphics. Some examples from Arizona:

Installation of a Data-Tracking Management System: The small neighboring communities of Pine and Strawberry sit astride the Mogollon Rim in central Arizona and draw most of their water from deep bedrock fracture zones. To better manage their limited water supply, the local citizen-volunteers of the Pine-Strawberry Water Improvement District (PSWID) invested in a bare-bones GIS system. PSWID purchased a desktop computer and GIS software, and developed a small, menu-driven database that allows updates of water levels and Arizona Department of Water Resources well registry data directly into the GIS

system. With this investment, PSWID can obtain and display the latest data and graphics from state and local agencies. More importantly, they have can track, update, and present local well and water level data, and thus make more informed management decisions for the community.

Baseline Data Compilation: Encroaching development on one of southern Arizona’s few perennial riparian areas

prompted Pima County to compile existing data covering six townships in the eastern portion of the county. The area was represented by GIS overlays of topography, orthophotos, stream channels, infrastructure, and land use. GIS was most useful for integrating the Arizona well registry database and water quality data over the base layers. Although much of the available data were of marginal quality, this inexpensive compilation produced a useful status map, revealing more than 1,000 wells in the area and showing close agreement with the known surface and subsurface hydrogeology of the area.

Urban Basin Storm Flow Evaluation: As part of a stormwater recharge feasibility study of a southern Arizona watershed, the Upper San Pedro Partnership used a GIS overlay of land use, soil type, vegetation, and topography to generate a runoff model of planes, channels, and impervious areas. The GIS-based Automated Geospatial Watershed Assessment (AGWA) tool (see next page) was then used to predict stormwater runoff peak flows, durations, and volumes from design storms and historical precipitation records. The model agreed well with other runoff estimation methods and was used to predict runoff from low-intensity storm events under pre- and post-development conditions.

In short, GIS also provides a wide variety of relatively low-cost and even low-tech applications on a local scale that can benefit multiple users, from agencies to community volunteers.

Contact Howard Grahn at howard@gsanalysis.com.

Not Just for Mega-Projects: Small-Scale GIS ApplicationsHoward Grahn – GeoSystems Analysis Inc.

May/June 2004 • Southwest Hydrology • 29

Dave Goodrich, Ph.D., Darius Semmens, Ph.D., and Averil Cate – USDA ARS

•AGWA (“Dot AGWA”) is the Web version of the Automated Geospatial Watershed Assessment tool developed by the USDA’s Agricultural Research Service, in cooperation with the Environmental Protection Agency and the University of Arizona. AGWA was developed as a multipurpose hydrologic analysis system for use by watershed, water resource, land use, and biological resource managers and scientists developing watershed and basin-scale studies. AGWA incorporates several spatial datasets, GIS mapping, analysis, and visualization tools, and surface water hydrologic models into one package, providing easy access to these features. The current version of AGWA is a desktop-based application driven by ESRI’s ArcView 3.x software (see tucson.ars.ag.gov/agwa/).

•AGWA is currently being developed at the USDA’s Southwest Watershed Research Center, also with the University of Arizona’s Watershed Resources Program and EPA. •AGWA takes the same features in AGWA and makes them available through a Web-based interface. •AGWA uses ESRI’s ArcIMS and Spatial Data Engine (SDE) and Oracle’s spatial database to provide the GIS data and interactions. •AGWA also uses Java-based Web server technology to connect the surface water hydrologic models to the application. The user is no longer restricted to the desktop version of the application, but can access the program remotely using a Web browser. The tool leverages client-server architecture so that changes and improvements in core components will not disrupt end-user interaction with the application. Because the application is Web-based, users also have immediate access to the latest versions of the software and GIS data. •AGWA will be available in late 2004 or early 2005.

For more information, contact David Goodrich at 602-437-1702 x241 or dgoodrich@uswcl.ars.ag.gov.

David L. Jordan, P.E. – INTERA Inc., and Peggy Barroll, Ph.D. – New Mexico Office of the State Engineer

A GIS technique was used to evaluate the magnitude and significance of phreatophyte evapotranspiration

(ET) from riparian areas along the Pecos River in southeastern New Mexico. The magnitude of ET was then compared with other water-balance components of the Carlsbad Area Groundwater Model (CAGW), which is used to evaluate management alternatives for the Pecos River. The CAGW is a MODFLOW model that simulates groundwater flow in the Capitan Reef and the overlying alluvial aquifer in the Pecos River Basin from Lake Avalon to the Malaga Bend. It simulates surface-water irrigation, groundwater diversions, transmission losses, and return flows. It also predicts base inflows to the Pecos River. Riparian ET is a potentially significant part of the water balance for the model, and thus must be evaluated and quantified.

A literature search was conducted to determine the rates of evaporation (EV) and ET for specified surface cover types in the CAGW model area. An ArcView shapefile of land cover types developed from aerial photography was used to determine riparian areas along the Pecos River and its tributaries. The cover types in the GIS coverage included cattail/

bullrush aspect, salt cedar, shrub wetlands (such as willow), and wet flats (mud flats with little or no vegetation). While the GIS coverage provides a good first estimate, additional field checking may be required. For example, areas designated “wet flats” may need to be reevaluated because wet conditions in these areas may occur only during certain flow regimes, or the areas may in fact be dry the majority of the time.

The EV/ET rates for each of the cover types were tabulated from available published literature sources. Annual ET rates were calculated from daily values given in the published sources. An attempt was made to extrapolate growing-season-length data over a one-year period. Most ET occurs during the growing season; therefore, data that were recorded only for the duration of the growing season were multiplied by 1.18 (100/85) to estimate annual rates. The annual EV and ET rates were then multiplied by the areal extent of each cover type (calculated using a GIS) to determine the water loss for each cover type within the CAGW active model layer one area. These results are tabulated below.

Contact David Jordan at djordan@intera.com or 505-246-1600 and Peggy Barroll at pbarroll@ose.state.nm.us or 505-827-6133.

Riparian Evapotranspiration Water Loss Estimates from the CAGW Model

Cover Type Acres1

ET Loss (acre-feet/yr) % of

Total ETLow High Average

Shrub Wetlands (Willows, etc.) 4 12 16 14 0.1%

Cattail/Bullrush 305 1,127 1,310 1,205 10%

Salt Cedar 824 1,281 19,907 5,449 44%

Wet Flats (mud) 1,824 3,030 8,333 5,682 46%

Totals 2,956 5,450 29,565 12,348 100%1Acreages calculated from GIS coverage

Evaluation of Potential Riparian Evapotranspiration Rates for a Groundwater Flow Model - A GIS Example from New Mexico

•AGWA GIS’s Future on the Web