As early as the 1890s, there had been talk of replacing the inefficient Wimmera Mallee open-channel system with a more-effective pipeline network that would provide a sustainable water supply to northwestern Victoria, Australia. But it was just talk--a pipe dream for the dryland farming and rural region extending from the Grampian Mountains in the south to the Murray River some 300 kilometers north.

Throughout the 20th century, despite its inefficiencies in supporting the area’s agriculture, the channel system was extended, branching out for a total of 17,500 kilometers (10,800 miles) in length. Meanwhile, the earthen channel system continued to draw water from already taxed sources such as the Wimmera and Murray rivers; reservoirs in the Grampians; and Victoria’s largest freshwater lake, Lake Hindmarsh. Add six years of below-average rainfall at the turn of the century, regional reservoirs at less than 10-percent capacity, and a highly dysfunctional channel system, and the result was a dryland region that could no longer sustain agriculture or prosperity.

Studies done in the late 1990s showed that of the 120,000 megaliters released into the channel system in an average year, only 17,000 or fewer were actually used by customers. The rest was lost to evaporation and leakage from the channels and farm dams.

By the year 2000, it became evident that the 100-year-old pipe dream needed to become a reality, and it had to happen quickly. Without a new pipeline system, the Wimmera Mallee region’s production of livestock, wheat, barley, canola, field peas, chickpeas and faba beans would cease to exist, not to mention the towns and businesses in the area, which required water to survive.

Members of the community and a steering committee heavily lobbied for a pipeline. Yet it was not until 2005 that the Commonwealth of Australia and the Victoria state government reached an agreement on the funding of the Wimmera Mallee Pipeline Project (WMPP). In the same year, it was announced that Grampians Wimmera Mallee Water (GWMWater) would be responsible for the delivery of the estimated 10-year project. But those with livelihoods at stake had to wonder if a completed pipeline in 10 years would be too late. 

An example of the 2-meter hillshade elevation grid created using ESRI ArcGIS data processing tools. Visible features include redundant on-farm storage dams as well as the open-channel network.

Let There Be LiDAR

Before GWMWater could start designing more than 5,592 miles of pipeline and building 40 pump stations, it had to undertake a time-intensive environmental analysis and survey of more than 4,530 square miles of land. The existing 10- and 20-meter contours and 1:25,000- and 1:50,000-scale line mapping available for most of the region was completely inadequate for this challenging and ambitious engineering project.

GWMWater knew that it would not be feasible to rely on photogrammetry and surveyors alone to capture and analyze the required data for such a large and diverse terrain. Additionally, the massive amount of resultant survey data needed to come in a format that would work with the ESRI GIS-based software platform that GWMWater had launched to manage the project. And every phase of the project had to move quickly. “The original 10-year construction plan was based on our experience with the Northern Mallee Pipeline Project, a much smaller pipeline, and on the proposed funding arrangements with state and federal governments,” says Jeff Rigby, GWMWater managing director. “However, we knew we needed to fast track the project in response to the drought conditions.”

An example of the 2-meter grid ALS thinned ground data, which was processed with ESRI ArcGIS 9.2.

The search for the right solution led GWMWater to AAMHatch, an Australian-based firm specializing in geospatial services and products. Rigby believed that significant efficiencies could be gained by using LiDAR-based imagery of the entire project area packaged in GIS spatial datasets, and AAMHatch possessed the LiDAR experience to make it happen. “The scale of the WMPP required ‘smart’ management of project data,” Rigby says. “LiDAR and GIS provided the opportunity for efficient sharing of data through complex planning processes, iterative design negotiations, tender and contract management.” 

According to Rohan Potter, AAMHatch geospatial consultant and project manager, the size of the firm’s aerial survey fleet and IT infrastructure combined with its skills in data processing, management and terrain product creation in the ESRI environment were the primary reasons AAMHatch was chosen for this time-critical project. “AAMHatch prides itself on delivering projects that are large and technically challenging,” Potter says. “This project certainly provided these opportunities!”

ArcGIS terrains provide optimized performance at multiple resolutions through the use of terrain pyramids that quickly retrieve only the data needed for the required level of detail in a given area of interest.

Covering New Ground

Vast volumes of data require careful data management and backup procedures to maximize efficient processing. Traditionally, elevation data have been supplied as an ASCII point file, an ESRI shapefile or a triangulated irregular network (TIN). However, these datasets are time consuming to create, load and analyze, and they have large storage requirements. For example, every 1 gigabyte ASCII file equates to a 2.5 gigabyte shapefile or TIN. Additionally, this process relies on a tile-based approach (generally comprising 2-by-2-kilometer tiles) that limits the number of points that can be used per file. And the file size recommendations are restricting--less than 3 million points per shapefile or 12 million points per TIN.

AAMHatch was determined to improve these capabilities for the WMPP. Prior to gathering the data, the firm worked closely with ESRI software developers to test and enhance the ArcGIS 9.2 software for large datasets. The solution was ESRI’s terrain dataset--a multiple-resolution TIN-based surface built from measurements stored as features in a geodatabase. ArcGIS terrains provide optimized performance at multiple resolutions through the use of terrain pyramids that quickly retrieve only the data needed for the required level of detail in a given area of interest. Having all these data in a GIS would prove crucial in project workflows and data sharing since the range of people accessing the WMPP GIS data varied from design engineers, planners, contractors and government agencies to auditors, project managers, field inspectors, land officers, and environmental and cultural heritage advisors. The resulting dataset would be seamless, and the file sizes would be easily manageable--a 1 gigabyte ASCII file or 200 megabyte Multipoint file becomes negligible in terrain.

Armed with multiple Optech LiDAR units and Intergraph large-format digital cameras, six two-person air crews in three aircraft worked day and night to capture the data and verify successful coverage. While the data-capture team gathered the GPS and airborne laser scanner (ALS) data, analysts processed LiDAR data for other geographic areas and generated derived GIS products along with terrain datasets of the project area comprising 1.3 billion points. Within 10 weeks of beginning the ALS data capture, AAMHatch was delivering digital terrain models (DTMs) of approximately +/-15 centimeters vertical accuracy at one sigma (standard deviation) on open areas with an average laser strike spacing of 1.3 meters--a significant improvement over the previously available data.

The data were delivered in more than six phases so that GWMWater and its contractors would receive components of the data as they became available. The terrain data were complemented by orthoimagery with 60-centimeter image resolution. Stored in the WMPP GIS, the LiDAR and orthoimagery data provided a current view of the land use, land cover, and terrain--all data that are essential for engineers working on the project. At the time, the spatial foundation dataset was the largest single-file terrain dataset ever produced in Australia.

The terrain data were easily and quickly processed and analyzed using the ArcGIS Server 3D extension. By generating terrain products and terrain files that were one-fifteenth the size of standard files, a savings of more than 20 days in data-processing time was achieved. “These capabilities significantly improved access, display, performance and analysis of LiDAR data and the elevation information these data represent,” Potter says.

The first pipes were laid in November 2006. Thanks to the use of LiDAR and advanced GIS capabilities, the WMPP is on schedule to be completed this year--five years earlier than originally projected. “There is no doubt that LiDAR technology enabled AAMHatch to deliver project data in reduced time,” says Bruce Van Every, PhD, GWMWater project director. “There were significant efficiencies gained through improved survey processes, accuracy of data, and the ability to share the data with a range of stakeholders through our GIS.”

When finished, the pipeline will supply stock and domestic water to 7,000 rural customers and 36 towns across the Wimmera Mallee region and save approximately 100 billion liters of water each year. “It is an enduring legacy of the project,” Rigby says, “a valuable resource for planning and development for the whole region.” 

Sidebar: Channeling the GIS Data Flow

The Wimmera Mallee Pipeline Project is one of the largest engineering initiatives currently under way in Australia. It spans an area of 11,700 square kilometers and aims to increase the efficiency of water distribution in the Wimmera and Mallee regions through pipelines and pump stations. GIS not only provided the required data for the design and construction phases of the WMPP but also was used to track and record ecological, soil, vegetation, and cultural heritage information; cadastral and landowner contact details; and lengths of pipes and valve types. More than 250 layers of GIS data were used with a total volume of approximately 1 terabyte.

GIS data continue to be a valuable resource for stakeholders. For example: