Figure 2. The Pit 3 dam is only one of four dams located within the project area (photo by author).

Remote Sensing for Hydroelectric Relicensing


Donald G. Price
Senior Scientist
Technical & Ecological Services Department
Pacific Gas and Electric Company
San Ramon, California

Pacific Gas and Electric Company (PG&E), a California Public Utility, operates extensive hydroelectric facilities on the Pit River of northern California , controlling a network of dams, tunnels, powerhouses and electrical transmission lines.

Figure 1. Location of the Pit 3, 4 and 5 hydroelectric project.
The Pit 3, 4, 5 Hydroelectric Project is a 325-megawatt hydroelectric facility located on the Pit River, in Shasta County , California (Figure 1.). The project occupies 746 acres of lands of the United States administered by the Forest Supervisors of the Shasta-Trinity and Lassen National Forests . The Pit 3, 4, 5 Project consists of three hydraulically connected developments, with a total of four dams, four reservoirs, three powerhouses, associated tunnels, surge chambers, and penstocks. The project has a combined average annual generation of 1,913.7 gigawatt-hours. The Pit 3, 4, 5 Project is operated by PG&E under Federal Energy Regulatory Commission (FERC) License No. 233.

The Pit River is a noted trout-fishing stream and home to diverse communities of wildlife and vegetation (Figure 2.). In 2002 a remote sensing program was conducted as one element of a major study of the potential effects of future hydroelectric power operations on the river ecosystem. This project and other studies were conducted as part of the FERC re-licensing process for the Pit 3, 4, 5 Project. Hydro re-licensing is a federally mandated process typically conducted once every 30 to 50 years for hydroelectric projects.

Figure 3. Instream flows are augmented by precise water releases from project dams like this 1,200 cfs test release from the Pit 3 dam (photo by author).

Figure 4. Three-meter resolution image mosaic of RGB bands produced by the HyMap hyperspectral flight over the entire 37 km length of the project area (image by HyVista Corporation).

Figure 5. Ground truth involved above-water and underwater evaluations with portable spectrometers and ground control point locations (photo by author).

Figure 6. Aquatic and terrestrial habitats were classified and mapped to determine baseline conditions and to estimate the effects of altered flow releases (image by Itres Inc.).
The process of hydroelectric generation impounds water behind dams and diverts it through tunnels and canals to lower elevation power stations (Figure 2). Water that would normally flow down the natural river channel may be reduced compared to pre-project conditions. This loss of flow in the natural river channel is mitigated by releasing very precise water flows from each of the dams situated along the river (Figure 3.). These flow releases are referred to as “instream flows” and their magnitude and timing are major issues.

A controlled flow study was conducted in the spring and summer of 2002 to evaluate the effects of various potential mitigation flow releases on aquatic resources, including assessments of fish habitat, fish stranding, amphibian habitat, mollusk habitat, filamentous algae movement, and sediment transport. Recreation concerns were also assessed with whitewater boating and fishability evaluations. Due to the diverse range of environmental studies requiring overhead imagery, several remote sensing methods were chosen to examine the full 37 km project length of the river encompassing all three power stations located within the Pit 3, Pit 4 and Pit 5 reaches.

In 2002 several remote sensing datasets were collected, including satellite panchromatic, terrestrial lidar, bathymetric lidar, airborne hyperspectral (Figure 4.), and seven acquisitions of high-resolution color airborne stereophotography. A ground truth dataset was collected to support each acquisition of remotely sensed data (Figure 5). All data were tied to the same horizontal and vertical datum for precise co-location between datasets in all three dimensions. The remote sensing datasets enabled mapping and modeling of the river and river basin morphology, classification of vegetation, and detection of several water quality parameters. Aquatic and terrestrial habitats were classified and mapped to determine baseline conditions and to estimate the effects of altered flow releases (Figure 6). These data were integrated into two-dimensional instream flow models of the river system, which simulate the spatial impact of different flow release rates and relate them to changes in aquatic habitat.

The results of these models are helping a collaborative team consisting of PG&E, public interest groups, and several regulatory agencies to determine specific mitigation flow release rates that approach an optimal compromise between habitat health and cost-effective power generation.

Remote sensing datasets collected in 2002, followed by Map Product:

Terrestrial Lidar , ± 15 cm vertical, 3 meter posting

a. Bare earth DEM
b. Canopy DEM
c. Test measurements of water surface elevations

Color Stereophotography, 10 cm resolution, 1:7200 scale

  • Precise river boundary vectors from stereo analysis
  • Base map for traditional ground surveys

Satellite Panchromatic Imagery

  • Watershed overview and terrain assessment
  • Planning for sampling locations and access

Hyperspectral Imagery, 1.5-meter and 3.0-meter, 126 spectral bands, 450 nm (nanometer) to 2480 nm

  • Water quality maps (chlorophyll content, total suspended solids)
  • Exposed and submerged substrate classification
  • River depth
  • Riparian vegetation classifications

Bathymetric Lidar, 4-meter posting

  • River bathymetry
  • Water surface elevation

Ground Surveys

  • GPS and total station positions
  • Spectral measurements
  • Water quality samples
  • Depth measurements
  • Site descriptions

List of Project Participants:

  • Technical & Ecological Services Department (TES) of Pacific Gas and Electric Company (PG&E, San Francisco, Calif.)
  • CSIRO Land and Water (Sydney, Australia )
  • DigitalGlobe Inc. (Longmont, Colo.)
  • EarthData International Inc. (Washington, D.C.)
  • HJW GeoSpatial Inc. (Oakland, Calif.)
  • HyVista Corporation (Sydney, Australia)
  • ITRES Limited (Calgary, Canada)
  • Joint Airborne Lidar Bathymetry Technical Center of Expertise (USACE JALBTCX, Mobile, Ala.)
  • Lawrence Livermore National Laboratory (LLNL, Livermore, Calif.)
  • University of California Santa Cruz (UCSC) Earth and Marine Sciences Department (Santa Cruz, Calif.)
  • Map Products

Figure 7. High resolution (0.6 meter) panchromatic imagery was used for initial study planning, for terrain assessment, and for locating key project features such as the Pit 4 powerhouse shown here (satellite image by DigitalGlobe Inc.).
Figure 8. High resolution (10 cm) color stereophotography was used for aquatic habitat mapping and for a variety of ground trut

Figure 9. Realistic 3-D views of the project area provide environmental scientists with a unique perspective of aquatic habitats. Lidar topographic maps were fused with vectors derived from hyperspectral data and high resolution aerial photos (image by Itres Inc.).
Figure 10. Hyperspectral imagery was classified into vectors that describe riparian vegetation, river depth, and benthic substrate types (image by Itres Inc.).
Satellite Panchromatic Imagery
was collected in the initial planning stages of the study to provide an overview of the river basin (Figure 7). QuickBird panchromatic imagery was used to identify specific stream segments that would overlap project facilities and to select ground control points for subsequent airborne activities.

High Resolution (10 cm) Color Stereophotography was collected seven times, once each at test flow releases from 150 cfs (cubic feet per second), 250 cfs, 400 cfs, 600 cfs, 800 cfs, 1200 cfs and 1800 cfs (Figure 8). The precise river boundaries at each river flow rate were digitized in stereo analysis to create 3D river boundary vectors accurate to the centimeter level. The stereophotography was also used as a basemap for a vegetation survey on the ground.

The lidar data were used to create a digital elevation model of the river canyon at a 3-meter spatial resolution. The lidar dataset contains classified returns of the bare earth, the vegetation canopy, the water and infrastructure. In addition, a successful test was conducted in the use of terrestrial lidar to map river water surface elevations (Figure 9).

Bathymetric lidar was used on an experimental basis to assess performance in a narrow, boulder-strewn river canyon. Not all the project area was mapped, but the valid returns agree with the elevation measurements from traditional lidar and ground surveys.

Hyperspectral imagery was used to map riparian vegetation, river depth and benthic substrate type (Figure 10). Spectral analyses and numeric models estimated the spatial concentrations of chlorophyll, total suspended solids and colored dissolved organic material in the Pit River .

Ground data collection to support this project included GPS and total station surveys, underwater and above-water spectral measurements, water depth measurements and water quality samples.

The application of remote sensing techniques for the environmental evaluation of hydroelectric projects is still developing. The value derived by using remote sensing will vary depending on the specific area of interest. Certainly, where access is difficult or impossible, the value of using remote sensing becomes extremely high. Aerial photography has been used for many decades for remote area mapping and remains an important remote sensing tool. New tools, such as high-resolution satellite imagery, airborne hyperspectral imagery, and lidar for topographic and benthic mapping are now available and enable many new ecosystem assessment approaches.

Remote sensing techniques are perceived as costly due to the high cost of aircraft, scanners, and data analysis, but actually compare favorably to more conventional techniques which are typically labor intensive. As remote sensing techniques improve, with better spatial and wavelength resolution, expensive on-the-ground fieldwork requirements will be reduced. Ultimately, however, it will be the unique point of view provided by remote sensing that will be the most valuable. Remote sensing not only allows a broad overview of landscapes of interest, it can also provide comprehensive analysis of critical areas that might be missed by conventional approaches.

Geospatial analyses of the variety of image data can produce a wealth of detailed specialized maps. Various layers can be superimposed and overlaid to show the distribution of plant species, the extent of aquatic habitats, and the juxtaposition of energy infrastructure in a compelling format.

Regulatory agencies, non-governmental organizations, and the general public have all responded favorably to the remote sensing products developed by this study, which has enhanced the environmental planning and decision making processes.


Pacific Gas and Electric Company (PG&E) makes no warranty or representation, expressed or implied, with respect to the accuracy, completeness, or usefulness of the information contained in this document, or that the use of any information, apparatus, method, or process disclosed in this document may not infringe upon privately owned rights. Nor does PG&E assume any liability with respect to use of, or damages resulting from the use of, any information, apparatus, method, or process disclosed in this document.

Sensors & Systems | Monitoring, Analyzing and Adapting to Global Change | Stay in tune with the transformation. Subscribe to the free weekly newsletter.