Mackenzie K. Keith
Joseph F. Mangano
2020
High-resolution orthophotograph of Fall Creek Lake, Oregon, acquired during annual drawdown to streambed November 10, 2016
raster digital data
data release
doi.org/10.5066/P9AYWU8Z
https://www.sciencebase.gov/catalog/item/5c79ade6e4b0fe48cb51450d
https://doi.org/10.5066/P9AYWU8Z
The Middle Fork Willamette River basin encompasses 3,548 square kilometers of western Oregon and drains to the mainstem Willamette River. Fall Creek basin encompasses 653 square kilometers and drains to the Middle Fork Willamette River. In cooperation with the U.S. Army Corps of Engineers, the U.S. Geological Survey evaluated geomorphic responses of downstream river corridors to annual drawdowns to streambed at Fall Creek Lake. This study of geomorphic change is focused on the major alluvial channel segments downstream of the U.S. Army Corps of Engineers’ dams on Fall Creek and the Middle Fork Willamette River, as well as the 736 hectare Fall Creek Lake. Reservoir erosion during streambed drawdown results in sediment delivery to downstream reaches. To better understand patterns and processes of reservoir erosion related to downstream geomorphic responses, detailed reservoir deposit mapping and change analyses are completed.
To support these analyses, aerial photographs were collected via unmanned aerial systems (UAS) on November 8 and November 9, 2016, and via Cessna aircraft (or manned aerial system; MAS) on November 10, 2016, to create 3-dimensional xyz point clouds, digital elevation models (DEMs), and orthophotographs of the Fall Creek Lake bottom during full reservoir drawdown using structure-from-motion (SfM) techniques. The November 8 model resulted in a ground resolution of 3.28 centimeters per pixel, average point density of 58.2 points per square meter, and DEM resolution of 13.1 centimeters per cell. The November 9 model resulted in a ground resolution on 2.86 centimeters per pixel, average point density of 76.6 points per square meter, and DEM resolution of 11.4 centimeters per cell. The model for the manned aerial system (MAS) flight on November 10 had a slightly lower ground resolution at 4.18 centimeters per pixel (for multiple altitudes), average point density of 35.8 points per square meter, and output DEM resolution of 16.7 centimeters per cell. Cell sizes were automatically generated by the software.
This documentation describes the high-resolution orthophotograph of Fall Creek Lake, Oregon, created from aerial photographs acquired when lake levels at streambed on November 10, 2016. This dataset is part of a broader publication including aerial photographs, 3-dimensional xyz point clouds, DEMs, orthophotographs, model processing reports, ground control points, and dataset footprints for November 8-10, 2016.
The raw photographs, point clouds, DEMs, and orthophotographs (herein collectively referred to as SfM datasets) covering Fall Creek Lake support detailed mapping of reservoir deposits and quantitative reservoir change analyses related to the streambed drawdown during water year 2017 (WY2017). The detailed mapping was primarily based on the November 10, 2016, SfM datasets and provide a basis for evaluating geomorphic context of eroded or deposited sediment relative to other (November 8 and 9) SfM datasets collected in WY2017, lidar of the reservoir collected in 2012, as well as other elevation datasets collected in the future. These datasets also provide basis for characterizing future geomorphic and topographic changes in Fall Creek Lake.
20161110
ground condition
None planned
-122.759073
-122.667780
43.973984
43.921350
ISO 19115 Topic Categories
inlandWaters
geoscientificInformation
imageryBaseMapsEarthCover
None
sediment transport
fluvial geomorphology
reservoir
aerial images
structure from motion
sfm
photogrammetry
unmanned aerial system
UAS
orthoimagery
orthophotograph
erosion
drawdown
dam
USGS Thesaurus
geospatial analysis
aerial photography
geomorphology
geospatial datasets
USGS Metadata Identifier
USGS:5dfd0b42e4b0b207aa00282e
Geographic Names Information System (GNIS)
Fall Creek
Winberry Creek
Middle Fork Willamette River
Oregon
Lane County
Fall Creek Lake
Jasper, Oregon
none
The U.S. Geological Survey shall not be held liable for improper or incorrect use of the data retrieved from the system. Public domain data from the U.S. Government are freely redistributable with proper metadata and source attribution. The U.S. Geological Survey should be acknowledged as the data source in products derived from these data.
U.S. Geological Survey - ScienceBase
Mailing and Physical
Denver Federal Center
Building 810
Mail Stop 302
Denver
CO
80225
United States
1-888-275-8747
sciencebase@usgs.gov
Browse orthomosaic November 10, 2016
High-resolution orthophotograph of Fall Creek Lake, Oregon, acquired during annual drawdown to streambed November 10, 2016
PNG
None
Unclassified
None
Version 6.2 (Build 9200); Esri ArcGIS 10.5.1.7333
Four UAS flights on November 8 captured 2,982 photographs covering 3.19 square kilometers at an average flight altitude (relative to the ground surface) of about 143 meters used to generate a dense point cloud in Agisoft PhotoScan Professional (version 1.2.6) software. A DEM and orthophotographs were also generated by the software. Similarly, two UAS flights on November 9 captured 1,702 photographs covering 1.72 square kilometers at an average flight altitude of 125 meters used to create point cloud, DEM, and orthophotographs. Photographs from UAS flights were captured with a Sony ILCE-5100 (6000x4000 resolution). UAS flights were contracted through the USGS to Frontier Precision, Inc., Boise, Idaho. The manned aerial survey on November 10 captured 1,888 photographs with a Ricoh GRII (4982x3264 resolution) over the study area but only 1,429 were used to generate a point cloud and subsequent DEM and orthophotograph. The photographs covered 6.78 square kilometers and were taken at different flight altitudes of about 180 and 300 meters. The manned aerial acquisition was contracted through the USGS to Brown Western Aviation, Independence, Oregon. Photographs were acquired by mounting the camera to the wing of the plane. All datasets are in UTM Zone 10 North, projection. The horizontal datum is the North American Datum of 1983 and elevations reference the North American Vertical Datum of 1988 (GEOID 12B).
All images obtained during the Fall Creek Lake drawdown during WY2017 are included in each associated zipped flight file. Photographs were separated by day to create point clouds, DEMs, and orthophotographs. Not all images were used in generation of SfM datasets due to various factors resulting in no-to-few tie points of ground features within overlapping images. Some manual removal of erroneous points was performed on the point cloud. All remaining points generated by the dense point clouds are included in the point cloud files for each day. Resulting DEMs and orthophotographs are not clipped to areas of high uncertainty, particularly near edges where minimal photograph overlap was used in photogrammetric processing. Additionally, DEMs have not been clipped to remove areas of vegetation that do not represent ground elevations.
No formal horizontal accuracy tests were performed. However, horizontal accuracy likely decreases with distance from ground control. Ground control points surveyed with a network-based RTK-GPS had a reported horizontal precision ranging from 0.01 to 0.05 meters.
Though, not an assessment of horizontal or vertical accuracy, residuals between the model fit and ground control points are described here. For the November 8 UAS model, thirty-two ground control point were used to construct SfM datasets, and the model had a horizontal error of 6.56 centimeters and vertical error of 3.51 centimeters (overall error of 7.44 centimeters). One ground control point was withheld and compared to the model and had a horizontal error of 2.31 centimeters, vertical error of 0.04 centimeters, and overall error of 2.31 centimeters. Because of the smaller area covered on the November 9 flights, only eighteen points were used to construct the model. Horizontal, vertical, and overall resulting model errors were 4.60, 4.60, and 6.51 centimeters, respectively. The one control point withheld had a horizontal error of 1.67 centimeters and a vertical error of 1.47 centimeters (overall error of 2.23 centimeters). For the November 10 MAS model, thirty-five ground control points were used to construct SfM datasets and had a horizontal error of 7.02 centimeters and vertical error of 3.44 centimeters (overall error of 7.82 centimeters). Four ground control points were withheld and compared to the model and had a horizontal error of 9.48 centimeters and vertical error of 7.18 (overall error of 11.89 centimeters). Visual comparison of DEMs generated for 2016 to lidar collected in 2012 indicated horizontal displacement was minimal for all datasets, although some obvious distortion is noticeable where vegetation is present or there was very little photograph overlap. Increasing noise at surface water locations occurs within each model; these areas were not edited or hydroflattened.
No formal vertical accuracy tests were performed. However, vertical accuracy likely decreases with distance from ground control. Ground control points surveyed with a network-based RTK-GPS had a reported vertical precision ranging from 0.01 to 0.05 meters.
Though, not an assessment of horizontal or vertical accuracy, residuals between the model fit and ground control points are described here. For the November 8 UAS model, thirty-two ground control point were used to construct SfM datasets, and the model had a horizontal error of 6.56 centimeters and vertical error of 3.51 centimeters (overall error of 7.44 centimeters). One ground control point was withheld and compared to the model and had a horizontal error of 2.31 centimeters, vertical error of 0.04 centimeters, and overall error of 2.31 centimeters. Because of the smaller area covered on the November 9 flights, only eighteen points were used to construct the model. Horizontal, vertical, and overall resulting model errors were 4.60, 4.60, and 6.51 centimeters, respectively. The one control point withheld had a horizontal error of 1.67 centimeters and a vertical error of 1.47 centimeters (overall error of 2.23 centimeters). For the November 10 MAS model, thirty-five ground control points were used to construct SfM datasets and had a horizontal error of 7.02 centimeters and vertical error of 3.44 centimeters (overall error of 7.82 centimeters). Four ground control points were withheld and compared to the model and had a horizontal error of 9.48 centimeters and vertical error of 7.18 (overall error of 11.89 centimeters). Comparison of the three DEMs generated for 2016 with random point sampling of the reservoir floor area showed an average vertical difference between November 8 and 9 was an average 0.07 meters, and the average vertical difference between November 8 and 10 was 0.09 meters; though, in both comparisons some values were much greater where there was notable uncertainty in the models. Vertical accuracy is poor around DEM edges, particularly farther from ground control. Increasing noise at surface water locations occurs within each model; these areas were not edited or hydroflattened.
Watershed Sciences, Inc.
2012
Lidar remote sensing data collection Fall Creek
remote sensing data
digital elevation model
20120113
20120223
ground condition
lidar 2012
The lidar provided a base layer from which horizontal and vertical accuracy were informally compared.
The USGS installed 40 temporary L-shaped control point targets throughout the Fall Creek Lake over November 7 and 8, 2016. The photo target controls mainly consisted of 1-foot wide white, high-reflectance Presco aerial flagging. The flagging was cut into 3-foot length strips. Two strips forming an L shape were nailed to the ground with 6-inch long nails. Additionally, two blue tarps were folded into an L shape and a handful of targets were constructed by nailing down two white trash bags in the form of an L.
20161107
Control point targets were surveyed with a network-based RTK-GPS at the inner point of the L on the target. The output coordinates were inspected against 2012 lidar and one point was removed because of poor vertical accuracy.
20161108
For aerial image acquisition with unmanned aerial systems, Frontier Precision, Inc. (Boise, Idaho), collected photographs with four separate flights on November 8, 2016 (11:37 to 16:00 Pacific Standard Time [PST]) and two separate flights on November 9, 1016 (10:49 to 12:21 PST). Photographs from UAS flights were collected with a Sony ILCE-5100 approximately every 1-2 seconds (.JPG extension). Four UAS flights on November 8 resulted in 2,982 photographs covering about 3.19 square kilometers at an average flight altitude (relative to the ground surface) of about 100 meters, while the two UAS flights on November 9 resulted in 1,702 photographs covering about 1.72 square kilometers also at an average flight altitude of about 100 meters.
20161108
For aerial image acquisition with the manned flight, Brown Western Aviation (Independence, Oregon) collected photographs on November 10, 2016 between 13:33 to 14:35 PST covering approximately 6.78 square kilometers of the reservoir area. A Ricoh GR II digital camera was mounted to the wing of Cessna aircraft and set to record RAW files (.DNG extension) every 2 seconds. The pilot flew a gridded pass at about 300 meters above ground level over Fall Creek Lake and then a repeated a longitudinal channel pass at about 180 meters above ground level over the main Fall Creek channel. The turns were slid instead of banked to keep camera level with the ground. A total of 1,888 photos were taken from takeoff to landing.
20161110
All 1,888 photographs from the November 10, 2016 flight were added to Agisoft PhotoScan software, and a medium-accuracy, preliminary alignment was used to find photos that were poor or would not align. The photoset was reduced to 1,463 photos that covered the main reservoir floor. Only 1,429 photos were used to generate dense point clouds, DEMs, and orthophotographs. For the UAS flights, all 2,982 photographs and all 1,702 photographs collected on November 8 and 9, respectively, were used for reconstruction. UAS camera position (GPS), yaw, pitch, and roll information was imported.
2017
For each day, photographs were aligned in with the following parameters: accuracy set to high; pair selection set to generic; key point limit and tie were left as the defaults; adaptive camera model fitting was checked.
2017
Control points were marked on all photographs where the control targets were visible.
2017
A control point file (.csv) was loaded with # Name (name of the control point), ym (y coordinate in meters), xm (x coordinate in meters), zm (z coordinate in meters), and accm (precision in meters, taken as the maximum error of the horizontal or vertical output from network-based RTK-GPS survey).
2017
The camera calibration model was optimized with the following coefficients: f, cx, cy, i1, k4, k3, k4, p1, p2, p3, p4, b1 and b2 where b1 or b2 are essentially the former aspect and skew coefficients; f is focal length; cx and cy are the optimal point; k is radial distortion; and p translational distortion.
2017
A dense point cloud was created using the medium-accuracy setting and aggressive filtering.
2017
A DEM was created using the dense point cloud and interpolation enabled. Alternative software could be used to produce improved DEM results from point cloud data.
2017
An orthomosaic was created using the Mosaic blending mode, DEM surface, and no color correction.
2017
Raster
Pixel
35000
35000
NAD 1983 UTM Zone 10N
0.9996
-123.0
0.0
500000.0
0.0
coordinate pair
0.000000002220024164500956
0.000000002220024164500956
meter
D North American 1983
GRS 1980
6378137.0
298.257222101
fc_day3_ortho
Pixel based raster with 3 layers (bands) of information for each pixel. Orthomosaic is split into multiple (9) tiles.
Other
Band_1
Red band.
Agisoft PhotoScan
0
232
Band_2
Green band.
Agisoft PhotoScan
0
241
Band_3
Blue band.
Agisoft PhotoScan
0
248
U.S. Geological Survey - ScienceBase
Mailing and Physical
Denver Federal Center
Building 810
Mail Stop 302
Denver
CO
80225
United States
1-888-275-8747
sciencebase@usgs.gov
Although these data have been used by the U.S. Geological Survey, U.S. Department of the Interior, no warranty expressed or implied is made by the U.S. Geological Survey as to the accuracy of the data. The act of distribution shall not constitute any such warranty, and no responsibility is assumed by the U.S. Geological Survey in the use of these data, software, or related materials. The use of firm, trade, or brand names in this report is for identification purposes only and does not constitute endorsement by the U.S. Geological Survey. The names mentioned in this document may be trademarks or registered trademarks of their respective trademark owners.
TIFF
ArcGIS 10.5
8-bit TIFF
8-bit unsigned integer TIFF with LZW compression, associated TIFF components, a browse thumbnail .png image, and FGDC CSDGM metadata. Total study area split into 9 tiles.
LZW compression.
11598
https://doi.org/10.5066/P9AYWU8Z
None. This dataset is provided by USGS as a public service.
20200827
Keith, Mackenzie K.
U.S. Geological Survey
Hydrologist
mailing and physical address
2130 SW 5th Avenue
Portland
Oregon
97201
United States
503-251-3474
mkeith@usgs.gov
FGDC Content Standard for Digital Geospatial Metadata
FGDC-STD-001-1998
local time