Hyperspectral and RGB image data used to map salmon spawning locations on the American River, California, November 5-7, 2018
Dates
Start Date
2018-11-05
End Date
2018-11-07
Publication Date
2020-07-02
Citation
Legleiter, C.J., and Harrison, L.R., 2020, Image data and field measurements used to map salmon spawning locations via remote sensing, American River, California, November 5-7, 2018: U.S. Geological Survey data release, https://doi.org/10.5066/P998CGA2.
Summary
An unmanned aerial system (UAS) was used to acquire red/green/blue (RGB) and hyperspectral image data from the American River in California November 5-7, 2018, to support research on remote sensing of rivers, specifically mapping chinook salmon spawning locations (redds) as part of an overall salmon habitat assessment program. The RGB images were acquired uisng a DJI MAtrice 210 equipped with a Zenmuse 4S optical camera. Imagery was collected along several parallel flight lines to ensure full coverage of the study reach. Raw Zenmuse images were georeferenced using real-time kinematic (RTK) global positioning system (GPS) and inertial motion unit (IMU) data recorded onboard the UAS and individual flight strips were combined into an [...]
Summary
An unmanned aerial system (UAS) was used to acquire red/green/blue (RGB) and hyperspectral image data from the American River in California November 5-7, 2018, to support research on remote sensing of rivers, specifically mapping chinook salmon spawning locations (redds) as part of an overall salmon habitat assessment program. The RGB images were acquired uisng a DJI MAtrice 210 equipped with a Zenmuse 4S optical camera. Imagery was collected along several parallel flight lines to ensure full coverage of the study reach. Raw Zenmuse images were georeferenced using real-time kinematic (RTK) global positioning system (GPS) and inertial motion unit (IMU) data recorded onboard the UAS and individual flight strips were combined into an orthorectified mosaic using Agisoft Metashape. The resulting orthoimage had a pixel size of 0.15 m. The georeferenced RGB ortho-image had accurate image coregistration with surveyed ground control targets used as checkpoints.
The hyperspectral data were collected by the Nano-Hyperspec imaging system, manufactured by Headwall Photonics, Inc. The flight was conducted by the NOAA Southwest Fisheries Science Center and colleagues from the University of California Santa Barbara who also performed the initial radiometric and geometric data processing. The data were acquired from a DJI Matrice M600 UAS and have a spatial resolution (pixel size) of 0.15 m. The data set consists of 252 spectral bands spanning the visible and near infrared wavelength range from 398 - 956 nanometers. Reflectance retrieval was performed using a calibration tarp as an in-scene white reference. The image pixel values represent reflectances and are stored as floating point 32-bit single precision numbers. The image data file has a band sequential (BSQ) interleave and is in an ENVI-compatible file format with an associated header (*.hdr) text file. The initial hyperspectral ortho-image had poor alignment with field-surveyed ground control points (GCPs) and required additional georeferencing to improve horizontal accuracy. We used ENVI software tools to perform image-to-image registration using the RGB ortho-image as the base image and the hyperspectral image as the warp image. Following this initial radiometric and geometric processing, the hyperspectral ortho-mosaic was masked to include only the water area within the river channel.
The images provided in this data release is focuses on the reach of the American River immediarely below Nimbus Dam. Supporting field data from this reach were collected in coordination with the acquisition of the remotely sensed data.
AmericanRgbWaterImage.hdr “Header file for water-only RGB image mosaic”
902 Bytes
text/plain
“Water-only hyperspectral image mosaic”
4.13 GB
application/unknown
AmericanRiverAtNimbusDam.JPG “Overview of redd mapping study site along the American River below Nimbus Dam”
127.07 KB
image/jpeg
AmericanImageListing.csv “Basic information about the images included in this data release.”
279 Bytes
text/csv
Related External Resources
Type: Related Primary Publication
Harrison, LR, Legleiter, CJ, Overstreet, BT, Bell, TW, Hannon, J. Assessing the potential for spectrally based remote sensing of salmon spawning locations. River Research and Applications. 2020; 36(8): 1618– 1632. https://doi.org/10.1002/rra.3690
For questions concerning this data set, please contact:
Dr. Carl J. Legleiter - cjl@usgs.gov, 303-271-3651
Geomorphology and Sediment Transport Laboratory
United States Geological Survey
4620 Technology Drive, Suite #400
Golden, CO 80403
Purpose
The purpose of this study was to develop and test a spectrally based technique for identifying salmon spawning locations, known as redds, from various types of remotely sensed data. Traditionally, redds have been mapped by eye while walking the bank or from a boat, or by an observer in an aircraft or an interpreter visually examining aerial images. The goal of this proof-of-concept investigation was to assess the potential for more efficient, objective, and automated redd mapping from conventional true color (RGB, or red/green/blue) and hyperspectral images.
Rights
Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.
Preview Image
Overview of redd mapping study site along the American River below Nimbus Dam