Geotagged sea-floor photographs and location of bottom photographs collected in Nantucket Sound, Massachusetts, in May 2016 and May 2017 by the U.S. Geological Survey during field activities 2016-005-FA and 2017-022-FA (JPEG images, point shapefile, and CSV file)

Step 1: Collected data.
Two marine geological surveys were conducted in Nantucket Sound, Massachusetts, in May 2016 and May 2017. The survey vessel occupied one of the target sites and the MiniSEABOSS was deployed off the vessel's starboard side in 2016 and off the vessel's stern in 2017. The MiniSEABOSS was equipped with a modified Van Veen grab sampler, two GoPro HERO4 Black digital cameras (the lower camera acquired bottom photos and the upper camera recorded backup bottom video), an oblique downward-looking SeaViewer 6000 HD Sea-Drop video camera with a topside feed, and a dive light to illuminate the sea floor for video and photograph collection. The elements of this particular MiniSEABOSS system were held within a stainless-steel frame that measured ~1 x 1 meter. The frame had a stabilizer fin that oriented the system as it drifted over the seabed. Two red lasers were set 20 centimeters apart (both as they are mounted on the MiniSEABOSS frame and as seen in photographs and video on the seabed) for scale measurements. The red laser dots can usually be seen in the sea-floor photos and videos depending on the bottom type and distance to the sea floor. A third laser is positioned at an angle so that when it intersects the other lasers, the MiniSEABOSS is at the optimum height (approximately 75 centimeters) off the bottom for a photograph. The winch operator lowered the sampler until the sea floor was observed in the topside live video feed. Generally, the vessel and sampler drifted with wind and current for up to a few minutes to ensure a decent photo with a clear view of the sea floor was acquired. The GoPro camera was set to capture photos every 2 to 5 seconds during a MiniSEABOSS deployment in 2016 and every 1 second in 2017. The photographed area is most often within 0.5 to 1.25 meters from left to right. Bottom video was also recorded during the drift. Then, the winch operator lowered the Van Veen sampler until it rested on the sea floor. When the system was raised, the Van Veen sampler closed and collected a sample as it was lifted off the sea floor. The sampler was recovered to the deck of the survey vessel where a subsample was taken for grain-size analysis at the sediment laboratory at the USGS Woods Hole Coastal and Marine Science Center. A total of 76 sites were occupied with the MiniSEABOSS: 55 sites were occupied aboard the R/V Rafael in May 2016 during field activity 2016-005-FA, and 21 sites were occupied aboard the R/V Tioga in May 2017 during field activity 2017-022-FA. Bottom photographs were collected at all 76 sites.

Step 2: Acquired and processed navigation.
During field activity 2016-005-FA, DGPS navigation from a Hemisphere DGPS receiver was logged through HYPACK navigation software and a DataBridge data logger. The DGPS was set to receive fixes at a 2-second interval in geographic coordinates (WGS 84). Dates and times were recorded in UTC. Log files for each MiniSEABOSS deployment were saved in text format and then combined for each Julian day. An AWK script (awkseth.gprmc.16005.awk) was used to parse the GPRMC navigation string from the log files and create ASCII Comma Separated Values (CSV) text files. The output files were merged and then reformatted using an AWK script (nav_time_reformat.awk), creating a processed navigation CSV text file for the survey (2016-005-FA_OdysseyAudioStamp_nav_ALL.csv). For field activity 2017-022-FA, a GPS log file in text format for the R/V Tioga was provided by the Woods Hole Oceanographic Institution. The navigation was collected at a 1-second interval in geographic coordinates (WGS 84), and dates and times were recorded in UTC. An AWK script (awksethTioga.gprmc.17022.awk) was used to parse the GPRMC navigation string from the log file, creating a processed navigation CSV text file for the survey (2017022FA_TiogaShip_nav.csv).

Step 3: Identified camera time offset for photos.
The original JPEG images do not represent spatial data; however, they were geotagged in a subsequent process step by linking the time the photo was taken, which was recorded in the JPEG files' Exchangeable image file format (Exif) metadata tags, to the corresponding time in the navigation files. Before geotagging the photos, the GoPro camera time, which was set to UTC, was compared to the GPS time using calibration photos to see if a time offset should be applied to the camera time of the photos. Camera time drift and the precision with which the camera time can be set often result in a time offset. Throughout the surveys, photos of an accurate UTC time (as shown on the navigation system or on the video, which had an overlay of the GPS time) were taken with the GoPro camera so that the camera time could be calibrated to the GPS time during photo processing. These calibration photographs indicated that the camera time was 22 seconds ahead of the GPS time on Julian day 123 and 29 seconds ahead of the GPS time on Julian day 124 during the 2016 survey. Calibration photographs were not available for Julian day 128 in 2016, so two photographs from that day were visually compared to the sea-floor video (which was recorded with an overlay of the GPS time) and found to have a camera time 46 seconds ahead of the GPS time. In 2017, calibration photographs indicated that the camera time was 92 seconds behind the GPS time. These time offsets were corrected in a later process step. Since different time offsets were applied to each survey day in 2016, the 2016 bottom photographs were renamed with a prefix of the Julian day (e.g., G0020192.JPG was renamed JD123_G0020192.JPG). This process step and the subsequent process steps were performed by the same person, Emily Huntley.

Step 4: Interpolated to create 1-second navigation.
The navigation fixes were interpolated to create 1-second navigation. This was done because the application used to geotag the bottom photos would have interpolated between fixes; therefore, before running the application, the navigation was interpolated so that the sediment and imagery data could be mapped using the same 1-second navigation. To interpolate the 2016 navigation data, first, a column of the original source filename was deleted from the processed navigation CSV text file (2016-005-FA_OdysseyAudioStamp_nav_ALL.csv). Next, a shapefile was created from the CSV file in Esri ArcGIS (version 10.3.1) and projected to WGS 84 UTM Zone 19N so that the distance in meters between the navigation fixes could be calculated. Then, the Generate Near Table tool was run using the planar method with the input and output features set to the UTM navigation shapefile and with the option checked to find only the closest feature. The output table was used to identify erroneous fixes, and two erroneous fixes were deleted from the navigation CSV text file. A Jupyter Notebook Python script (Interp_NAV_Dec2019_for_Nantucket2016-005-FA.ipynb) was run to interpolate and create a CSV text file of 1-second navigation. The original 2-second fixes were logged only when the video was recorded, so the script also interpolated between the end point of a video trackline and the start point of the next trackline. To create a final file of 1-second navigation, the points between the video trackline start and end times were extracted (which removed the interpolated positions during the transits between the video tracklines) and were saved as a CSV text file (out_interp_2016Nantucket_sel_for_gpx.csv). The CSV file was reorganized and formatted to have fields of the latitude, longitude, hours, minutes, seconds, Julian day, year, field activity ID, and Julian day and time. Finally, the CSV text file was converted to a GPX file (out_interp_2016Nantucket_sel.gpx), which is a GPS track log file format supported by ExifTool, using GPS Visualizer (http://www.gpsvisualizer.com/convert_input?convert_format=gpx, accessed December 20, 2019). For the 2017 survey, navigation data were logged for the full Julian day, including when the survey vessel was at the dock and when it was transiting to, from, and between the sampling sites. First, the navigation data from 13:30:00 to 19:14:59 were extracted from the processed navigation CSV text file (2017022FA_TiogaShip_nav.csv) to remove the navigation logged at, departing, and returning to the dock. Next, columns of the original source filename and time (the hours, minutes, and seconds were listed individually in other columns) were deleted from the CSV text file, and leading spaces were deleted as needed. Then, a shapefile was created from the CSV file in Esri ArcGIS (version 10.3.1) and projected to WGS 84 UTM Zone 19N so that the distance in meters between the navigation fixes could be calculated. The Generate Near Table tool was run using the planar method with the input and output features set to the UTM navigation shapefile and with the option checked to find only the closest feature. The output table was used to identify erroneous fixes, but no erroneous fixes were identified. The navigation data, however, were noisy. To smooth the data, the coordinates were rounded from seven to six decimal places and every fourth fix was extracted. A Jupyter Notebook Python script (Interp_NAV_Dec2019_for_Nantucket2017-022-FA.ipynb) was run to interpolate and create a CSV text file of 1-second navigation (2017022FA_TiogaShip_nav_4secint_round6_interp_for_gpx.csv). The CSV file was reorganized and formatted to have fields of the latitude, longitude, hours, minutes, seconds, Julian day, year, field activity ID, and Julian day and time to create a final CSV file of 1-second navigation for the survey. Finally, the CSV text file was converted to a GPX file (2017022FA_TiogaShip_nav_4secint_round6_interp.gpx) using GPS Visualizer (http://www.gpsvisualizer.com/convert_input?convert_format=gpx, accessed December 15, 2019).

Step 5: Geotagged the photos and incorporated survey information into the photos' metadata tags.
The following commands were used with ExifTool (version 11.86) to geotag the photos by populating the GPS tags and to incorporate additional survey information into the appropriate metadata tags of each JPEG image. Since the photos were acquired during two surveys and various time offsets were applied to each Julian day of surveying, different commands were run for each day's photos using slightly different tags. In the commands below, the "-geosync" value is equal to the time offset identified in a previous process step; it is the time difference in seconds that is added to the camera time before determining the photo's GPS time and location with a leading "+" or "-" added for positive or negative differences (negative if the camera clock was ahead of the GPS time). Survey information is duplicated in some tags because various software packages access different tags. Please note that depending on the software used to view the metadata tags, only a subset of these tags may be extracted. For more information about the ExifTool commands, see https://exiftool.org/geotag.html and https://exiftool.org/faq.html.
2016 photos from Julian day 123:
exiftool -api GeoMaxIntSecs=0 -api GeoMaxExtSecs=0 -P -m -geotag ./out_interp_2016Nantucket_sel.gpx -geosync=-22 '-geotime<${DateTimeOriginal}+00:00' -Artist="Laura L. Brothers" -Copyright="Public Domain - please credit U.S. Geological Survey" -ImageDescription="Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2016-005-FA" -Caption-Abstract="Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2016-005-FA" -Contact="WHSC_data_contact@usgs.gov" -CopyrightNotice="Public Domain - please credit U.S. Geological Survey" -Credit="U.S. Geological Survey" -sep ", " -keywords="Nantucket Sound, Massachusetts, 2016-005-FA, SEABOSS, sea floor, USGS" -comment="Photo from down-looking camera on the USGS SEABOSS deployed from the R/V Rafael during survey 2016-005-FA (https://cmgds.marine.usgs.gov/fan_info.php?fan=2016-005-FA)." -Caption="Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2016-005-FA" ./2016_originals/JD123*
2016 photos from Julian day 124:
exiftool -api GeoMaxIntSecs=0 -api GeoMaxExtSecs=0 -P -m -geotag ./out_interp_2016Nantucket_sel.gpx -geosync=-29 '-geotime<${DateTimeOriginal}+00:00' -Artist="Laura L. Brothers" -Copyright="Public Domain - please credit U.S. Geological Survey" -ImageDescription="Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2016-005-FA" -Caption-Abstract="Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2016-005-FA" -Contact="WHSC_data_contact@usgs.gov" -CopyrightNotice="Public Domain - please credit U.S. Geological Survey" -Credit="U.S. Geological Survey" -sep ", " -keywords="Nantucket Sound, Massachusetts, 2016-005-FA, SEABOSS, sea floor, USGS" -comment="Photo from down-looking camera on the USGS SEABOSS deployed from the R/V Rafael during survey 2016-005-FA (https://cmgds.marine.usgs.gov/fan_info.php?fan=2016-005-FA)." -Caption="Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2016-005-FA" ./2016_originals/JD124*
2016 photos from Julian day 128:
exiftool -api GeoMaxIntSecs=0 -api GeoMaxExtSecs=0 -P -m -geotag ./out_interp_2016Nantucket_sel.gpx -geosync=-46 '-geotime<${DateTimeOriginal}+00:00' -Artist="Laura L. Brothers" -Copyright="Public Domain - please credit U.S. Geological Survey" -ImageDescription="Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2016-005-FA" -Caption-Abstract="Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2016-005-FA" -Contact="WHSC_data_contact@usgs.gov" -CopyrightNotice="Public Domain - please credit U.S. Geological Survey" -Credit="U.S. Geological Survey" -sep ", " -keywords="Nantucket Sound, Massachusetts, 2016-005-FA, SEABOSS, sea floor, USGS" -comment="Photo from down-looking camera on the USGS SEABOSS deployed from the R/V Rafael during survey 2016-005-FA (https://cmgds.marine.usgs.gov/fan_info.php?fan=2016-005-FA)." -Caption="Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2016-005-FA" ./2016_originals/JD128*
2017 photos:
exiftool -api GeoMaxIntSecs=0 -api GeoMaxExtSecs=0 -P -m -geotag ./2017022FA_TiogaShip_nav_4secint_round6_interp.gpx -geosync=+92 '-geotime<${DateTimeOriginal}+00:00' -Artist="Laura L. Brothers" -Copyright="Public Domain - please credit U.S. Geological Survey" -ImageDescription="Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2017-022-FA" -Caption-Abstract="Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2017-022-FA" -Contact="WHSC_data_contact@usgs.gov" -CopyrightNotice="Public Domain - please credit U.S. Geological Survey" -Credit="U.S. Geological Survey" -sep ", " -keywords="Nantucket Sound, Massachusetts, 2017-022-FA, SEABOSS, sea floor, USGS" -comment="Photo from down-looking camera on the USGS SEABOSS deployed from the R/V Tioga during survey 2017-022-FA (https://cmgds.marine.usgs.gov/fan_info.php?fan=2017-022-FA)." -Caption="Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2017-022-FA" ./2017_originals
The following tags were populated in the JPEG image headers using the above ExifTool commands:
GPS tags (these values are unique for each photo and based on the matched-up location data):
GPSLatitudeRef GPSLatitude GPSLongitudeRef GPSLongitude GPSTimeStamp GSPDateStamp
Exif tags and the information used to populate these tags:
Artist: Laura L. Brothers
Copyright: Public Domain - please credit U.S. Geological Survey
ImageDescription (2016 photos): Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2016-005-FA
ImageDescription (2017 photos): Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2017-022-FA
International Press Telecommunications Council (IPTC) tags and the information used to populate these tags:
Caption-Abstract (2016 photos): Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2016-005-FA
Caption-Abstract (2017 photos): Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2017-022-FA
Contact: WHSC_data_contact@usgs.gov
CopyrightNotice: Public Domain - please credit U.S. Geological Survey
Credit: U.S. Geological Survey
Keywords (2016 photos): Nantucket Sound, Massachusetts, 2016-005-FA, SEABOSS, sea floor, USGS
Keywords (2017 photos): Nantucket Sound, Massachusetts, 2017-022-FA, SEABOSS, sea floor, USGS
JPEG tag and the information used to populate this tag:
Comment (2016 photos): Photo from down-looking camera on the USGS SEABOSS deployed from the R/V Rafael during survey 2016-005-FA (https://cmgds.marine.usgs.gov/fan_info.php?fan=2016-005-FA).
Comment (2017 photos): Photo from down-looking camera on the USGS SEABOSS deployed from the R/V Tioga during survey 2017-022-FA (https://cmgds.marine.usgs.gov/fan_info.php?fan=2017-022-FA).
Extensible Metadata Platform (XMP) tag and the information used to populate this tag:
Caption (2016 photos): Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2016-005-FA
Caption (2017 photos): Photograph of the sea floor in Nantucket Sound, Massachusetts, from survey 2017-022-FA
To extract the information from the photo metadata tags using ExifTool, the following command can be used (tested with ExifTool version 11.86):
exiftool -csv -f -filename -GPSTimeStamp -GPSDateStamp -GPSLatitude -GPSLongitude -c "%+.7f" -Artist -Copyright -ImageDescription -Caption-Abstract -Contact -CopyrightNotice -Credit -keywords -comment -Caption *.JPG > out.csv
The -csv option writes the information to a CSV file. The -c option formats the latitude and longitude as signed decimal degrees with seven decimal places. Please note that the above command was tested on macOS; Windows users may need to use "exiftool.exe" instead of "exiftool."

Step 6: Renamed the photos.
The photographs were renamed to include the field activity identifier, camera, and GPS date and time in the ISO 8601 standard (YYYYMMDD T [time separator] HHMMSS Z [Zulu/UTC time]) in the filename. Renaming the photos was done by creating a shell script with the original photo name and the new photo name with the GPS date and time information parsed from the photo's Exif metadata tags.

Step 7: Culled the photos.
Photos with no corresponding navigation data, which included 118 photos from the 2016 survey and 99 photos from 2017, were removed from the dataset. Most of these photos were taken on deck or in the water column. Since the photos were captured every 2 to 5 seconds in 2016 and every 1 second in 2017, the remaining bottom photographs were culled to select a few representative photos for each site for publication. To do this, photos that intersected a bottom video trackline were selected to remove photos collected within the water column or on deck. The selected photos were then reviewed to choose representative photos with a clear view of the sea floor. If a site had many photos with a clear view of the sea floor, then a photo generally every 15 to 20 seconds was selected for publication. Navigation was not recorded for site 2016-005-043, so a bottom photograph taken when the MiniSEABOSS was resting on the sea floor was given the coordinates of the closest available fix, which was shortly after the MiniSEABOSS was lifted off the sea floor at the site, using ExifTool and added to the dataset. An additional photo that did not intersect a video trackline was also added to the dataset; one photo that was taken a few seconds before the start of the video trackline at site 2016-005-044 was selected for publication since it had a clear view of the sea floor.

Step 8: Created a CSV file of the bottom photo locations.
The information from the photo metadata tags was extracted using the following ExifTool (version 11.86) command:
exiftool -csv -f -filename -GPSTimeStamp -DateTimeOriginal -GPSDateStamp -GPSLatitude -GPSLongitude -c "%+.7f" *.JPG > out.csv
Microsoft Excel 2016 for Mac was used to rename the field names for the filename (PICNAME), GPS time (GPSTIME), original camera time (CAMR_TIME), GPS date (DATE), latitude (LAT), and longitude (LONG); add new fields including an attribute for the camera (CAMERA), time offset applied to the camera time (TIMEOFFSET), Julian day of collection (JD), year of collection (YEAR), survey ID (FA_ID), sampling device used to collect the photo (DEVICE_ID), and survey vessel (VEHICLE_ID); and remove additional fields extracted from the ExifTool command.

Step 9: Created a final shapefile and CSV file of the bottom photo locations.
A shapefile was created from the bottom photos locations CSV file using the XY Table To Point tool in Esri ArcGIS Pro (version 2.4). The Spatial Join tool was then run to calculate the site number (FIELD_NO) and video trackline (LINENAME) for each bottom photo by joining the attributes of the bottom video tracklines shapefile using the intersect match option with a search radius of 0.01 meters. Two of the bottom photos did not intersect a bottom video trackline, so the site number (FIELD_NO) was assigned to the photo and a no data value of -9999 was added to the blank video trackline attribute (LINENAME). Unnecessary fields created when running the Spatial Join tool were deleted (i.e., Join_Count and TARGET_FID). Finally, the shapefile's attribute table was exported in Esri ArcGIS Pro and saved as a CSV file. The FID field was deleted in the CSV file.