During March 1- March 12 2022, July 20-28 2022, February 16-March 1 2023 and August 10-20 2023 an underwater imager (In Situ Ichthyoplankton Imaging System 3, ISIIS-3) was deployed. Winter (March 2022 and February-March 2023) deployments were from the R/V Sikuliaq, while summer 2022 (July) deployments were from the R/V Marcus G. Langseth, and summer 2023 (August) deployments were from the R/V Sally Ride. The ISIIS-3 deployments were carried out from these R/Vs in the northern California Current by a team affiliated with Oregon State University, the University of Oregon, and Portland State University. The ISIIS-3 was equipped with two line scan cameras for collecting plankton imagery in addition to six environmental sensors. This dataset contains the processed and quality controlled merged environmental data from the ISIIS imager on these cruises, consisting of temperature, conductivity, pressure, depth, salinity, fluorescence, oxygen, pH, photosynthetically active radiation, and chlorophyll a. Latitude, longitude, date, time, imager horizontal/vertical speeds, and altitude are also included in the dataset. Data were collected on six cross-shelf transects ranging geographically from the border of California and Oregon to southern Washington. In winter 2023, only five transects were sampled due to sampling constraints associated with winter storms. Two transects, one following the Newport Hydrographic Line and the other near Gray’s Harbor, Washington, were 70 nautical miles (~135 km) in length. The four others (near the Columbia River, Cape Meares, Heceta Bank, and the Rogue River) were 40 nautical miles (~73 km). ISIIS was towed continuously in an undulating motion (“tow-yos”) between 0-100m depth, or within 2-5m of the seafloor where the bottom depth was less than 100m. Data were streamed on to the ship in real time using a fiber optic oceanographic cable and written into millisecond time-stamped data files for each sensor. Sensor data files from ISIIS-3, along with latitude and longitude data from ship data streams, were merged using Python scripts into a final dataset based on the millisecond time stamps of the sensors and GPS streams.

Table of Contents

• Coverage
• Dataset Description
  • Methods & Sampling
  • Data Processing Description
  • BCO-DMO Curation Notes
  • Problem Description
• Data Files
• Supplemental Files
• Related Publications
• Parameters
• Instruments
• Deployments
• Project Information
• Funding

Location description: Six cross-shelf transects in the northern California Current, ranging from near the Oregon-California border (42.5°N) to southern Washington (47.1°N). Maximum depth of instrument is 100m or within 2-5m of the seafloor. A list of more detailed information about the individual transects is included.

ISIIS-3 was deployed from the research vessels via a fiber optic enabled winch towing system, and towed the length of the transect in an undulating manner at 2.5 m/s, undulating between 1m and 100m depth, or within 2m of the seafloor in waters shallower than 100m depth. A total of six transects (four in February 2023) were sampled, all during nighttime hours, with the longest transects split into inshore and offshore portions sampled over two nights. Each transect was sampled in the “tow-yo” fashion throughout the deployment of the instrument. Data were streamed to the shipboard ISIIS-3 computer in real time via a fiber optic connection through the winch. A program manager written in JADE took the input data streams from both cameras, each sensor, and GPS data from the ship, and wrote into individual millisecond time-stamped files. Data were stored on the shipboard computers on two 16TB drives, then copied on to individual 6TB external drives in duplicate for each camera. Individual files were written for imagery, CTD, GPS stream, voltage-based sensor readings (pH, dissolved oxygen, PAR), altimeter, and flowmeter.

See Schmid et al. (2023, doi:10.3389/fmars.2023.1187771) for more information on ISIIS-3 imager and deployments on these research cruises. 

Sensor data were written by individual publishers and millisecond time stamped. A Python script was used to merge these individual files into a final dataset based on the millisecond time stamps associated with each observation, creating a cohesive dataset over the time period sampled. Conversion factors and calibrations were applied to pH, oxygen, and irradiance voltage readings to yield final values, and chlorophyll-a was calculated using manufacturer conversions and calibration values. Conversion factors and calibrations were applied in R. 

Histograms of sensor data values were used to determine extreme values to flag as potential outliers. 

This section documents curation actions performed prior to publication review with the submitter, and additional information relevant to understanding and reusing this dataset. It distinguishes changes made to the submitted (meta)data from unresolved issues and/or enhancements that improve future reuse and interoperability.

CURATION ACTIONS PERFORMED ON DATA

- Loaded four source files: W2023_ISIIS_environmental_data_BCODMO.csv, W2022_ISIIS_environmental_data_BCODMO.csv, S2022_ISIIS_environmental_data_BCODMO.csv, S2023_ISIIS_environmental_data_BCODMO.csv, treated as csv format with headers in row 1, cast as strings, missing values "", "nd", "NA" defined
- Concatenated the four loaded tables into single table named 942686_v2_isiis-environmental-data, mapping matching columns, added file_name column recording source file name
- Applied find/replace on Time column fixing two specific timestamp strings (2023-08-14 02:36:48 and 2022-03-06 04:13:33) to add missing microsecond precision (.000000)
- Converted Time column (parsed as %Y-%m-%d %H:%M:%S.%f, timezone US/Pacific, confirmed by data submitters as PDT/PST) into ISO_DateTime_UTC column in UTC, formatted as %Y-%m-%dT%H:%M:%S.%fZ
- Reordered columns to: Time, Cruise_ID, ISO_DateTime_UTC, Temperature, Conductivity, Pressure, Depth, Salinity, Altitude, Horizontal_Speed, Vertical_Speed, Fluoro, Latitude, Longitude, PAR, pH, Oxygen, chl_a_ul, Temperature_flag, Conductivity_flag, Pressure_flag, Depth_flag, Salinity_flag, Horizontal_Speed_flag, Vertical_Speed_flag, Fluoro_flag, PAR_flag, pH_flag, Oxygen_flag, chl_a_flag, flag_totals, file_name
- Deleted file_name column
- Rounded Latitude, Longitude, PAR, pH, Oxygen columns to 5 decimal places, preserving trailing zeros
- Output file as 942686_v2_isiis_environmental_data.csv as primary file for version 2

CURATION ACTIONS PERFORMED ON METADATA

- BCO-DMO's standard metadata entry and text formatting steps were performed. See: https://www.bco-dmo.org/how-to/standard-curation-edits
- Adjusted metadata to include new cruises

ISSUES POTENTIALLY IMPACTING REUSE

- N/A

VERSION NOTES

Version 2:
Version date:
- This dataset has be updated to include additional data from deployments from the winter of 2022 and winter of 2023, cruises SKQ202204S and SKQ202303S
- The summary files no longer include "Transect_length_km" and "Transect_length_nm" columns

Note: This section refers to "NA" values to indicate missing data. Missing data values at BCO-DMO vary depending upon the format you download. For example blanks (null) are provided in csv format, NaN in Matlab files, etc.


[Problems/Issues for Summer 2022 Deployment]

NA values in the dataset correspond to missing data values. There are known gaps from 2022-07-24 07:17:06-07:17:45, 07:14:06-07:14:59, and 07:28:04-07:28:21 where most sensors read NA. There are several additional single second gaps that similarly show NA values for many sensors. These are flagged in the flag columns. These data gaps are likely due to cable connection movement on the ISIIS-3 imager where data were unable to be transferred from the electronics pod to the oceanographic cable.

Flag columns were created for each of the sensor values based on histograms of the values to determine potential outliers or unreliable values. These are described below, plus in the parameter description file.

Depth_flag: values less than 1m, greater than 110m, or NA flagged as 1. Values between 1-110m are 0.

Pressure_flag: values less than 1dBar, greater than 110 dBar, or NA flagged as 1. Values between 1-110dBar are 0.

Salinity_flag: values less than 20 PSU or NA flagged as 1. Values greater than or equal to 20 PSU are 0.

Conductivity_flag: values <3 S/m or NA flagged as 1. Values greater than 3 S/m are 0.

Temperature flag: NA values are flagged as 1. Non-missing data are 0.

Horizontal_Speed_flag: values greater than 500 cm/s, less than -500 cm/s, or NA flagged as 1. Values between -500 and 500 cm/s are 0.

Vertical_Speed_flag: values <0 cm/s,  greater than 4000 cm/s or NA flagged as 1. Values between 0-4000 cm/s are 0.

Fluoro_flag: values >4000 counts, <0 counts, or NA flagged as 1. Values between 0-4000 are 0.

PAR_flag: values < 0 µE/s, >0.1 µE/s, or NA flagged as 1. Values between 0-0.1 µE/s are 0.

pH_flag: values <7 or NA flagged as 1. Values ≥ 7 are 0.

Oxygen_flag: NA values flagged as 1, otherwise 0.

Chl_a_flag: values <0 µl/l, > 45 µl/l, or NA flagged as 1. Values between 0 and 45 µl/l are 0.

Flag_totals: total sum of flags for other columns. Ranges between 0 (no data flags) and 12 (data to be scrutinized or many extreme values)



[Problems/Issues for Summer 2023 Deployment]

The oxygen sensor and pump attached to the ISIIS-3 malfunctioned during all deployments. Therefore, the data for oxygen should not be used, and the quality control column is flagged for every value.

Occasionally, short periods of time do not have associated sensor values due to short data transfer lapses. In these cases, many of the instrumentation values read as NA, and

Flag columns were created for each of the sensor values based on histograms of the values to determine potential outliers or unreliable values. These are described below, plus in the parameter description file.

Depth_flag: values less than 1m, greater than 110m, or NA flagged as 1. Values between 1-110m are 0.

Pressure_flag: values less than 1dBar, greater than 110 dBar, or NA flagged as 1. Values between 1-110dBar are 0.

Salinity_flag: values less than 20 PSU or NA flagged as 1. Values greater than or equal to 20 PSU are 0.

Conductivity_flag: values <3 S/m or NA flagged as 1. Values greater than 3 S/m are 0.

Temperature flag: NA values are flagged as 1. Non-missing data are 0.

Horizontal_Speed_flag: values greater than 500 cm/s, less than -500 cm/s, or NA flagged as 1. Values between -500 and 500 cm/s are 0.

Vertical_Speed_flag: values <0 cm/s,  greater than 4000 cm/s or NA flagged as 1. Values between 0-4000 cm/s are 0.

Fluoro_flag: values >4000 counts, <0 counts, or NA flagged as 1. Values between 0-4000 are 0.

PAR_flag: values < 0 µE/s, >0.1 µE/s, or NA flagged as 1. Values between 0-0.1 µE/s are 0.

pH_flag: values <7 or NA flagged as 1. Values ≥ 7 are 0.

Oxygen_flag: NA values flagged as 1, otherwise 0.

Chl_a_flag: values <0 µl/l, > 45 µl/l, or NA flagged as 1. Values between 0 and 45 µl/l are 0.

Flag_totals: total sum of flags for other columns. Ranges between 0 (no data flags) and 12 (data to be scrutinized or many extreme values)

Website: http://hmsc.oregonstate.edu/research-labs/planktonlab/current-research

Coverage: Northern California Current

NSF Award Abstract:
Marine plankton form the base of most ocean food webs that support valuable fisheries. This highly diverse and complex community is composed of organisms that drift with ocean currents. Planktonic organisms remain understudied: they are difficult to sample given that their sizes span more than six orders of magnitude from less than one micron to meters. Yet, understanding how these communities respond to climate change, and ultimately how these responses affect valuable fisheries, and therefore food security, is critical. Because many ecological and physiological processes are dictated by relative size, the theory of size spectra (i.e., the relationship between size and organism abundance as it drives ecosystem properties such as food webs) provides a valuable framework for forecasting climate change impacts on marine ecosystems. A deeper understanding of the scope and nature of variability in size spectra under contrasting environmental conditions is needed. The dynamic, highly productive northern California Current off Oregon and Washington, during the summer and winter seasons, produces a patchwork of oceanographic conditions including those associated with hypoxia and ocean acidification. This study is sampling the plankton communities in this region to investigate how gradients of temperature, nutrients, dissolved oxygen, and pH conditions impact size spectra. The broader impacts include the training of students, building scientific resources, and outreach to broader communities. Undergraduate and graduate students are being trained in oceanography, field research and new technologies. The automated image analysis pipeline developed as part of the project is openly accessible to the oceanographic community and the image data are available through the novel Global Plankton Imagery Library, an open-access repository for plankton imagery. Size spectra data from this study are shared directly with ecosystem modelers. The project’s flagship outreach activity is the collaboration with the Sitka Center for Art and Ecology and the hosting of an Artist-At-Sea Program. A professional artist is competitively selected to join the research cruises and to create artistic products that give a unique voice to oceanographic research and the organisms under study. The artwork is being assembled into a traveling public Art Exhibit with planned displays at the Sitka Center, Oregon State University’s Hatfield Marine Science Center, University of Oregon’s Charleston Marine Life Center and centers located in underserved coastal communities. Finally, imagery data from the project are being shared via the Plankton Portal, a public website developed in partnership with the Citizen Science Alliance’s Zooniverse, that invites citizen scientists to participate in classifying plankton images.

The coupling of in situ plankton imagery and morphometric data allows quantifying scales of variation in plankton size spectra as well as testing predictions of how changes in environmental conditions (notably, temperature, nutrients, oxygen, pH) correlate with shifts in size spectra to reveal functional consequences to the food web. Plankton size spectra are being compared across environmental conditions by sampling in a habitat with steep environmental gradients and during two contrasting seasons. Planktonic organisms spanning 10 orders of magnitude in biomass are sampled using two complementary high-resolution imaging systems: the In Situ Ichthyoplankton Imaging System (ISIIS) and the Laser In-Situ Scattering and Transmissometry (LISST) particle imager. High-throughput image analysis software is used to create size distributions together with taxonomic classification. Depth-discrete meso-zooplankton samples are collected in parallel to examine community shifts in carbon, obtain length-to-carbon conversions and calibrate image data. The normalized biomass size spectra computed from the image data are tested for deviations from expected patterns. The plankton collections are also being analyzed for diet and reproductive status of gelatinous zooplankton, and diet and daily growth rate of representative larval fishes. These two groups have been historically understudied yet play central roles in ecosystem function. The data are being used to examine how these organisms are impacted by environmental conditions, and how they affect plankton size spectra. This study is foundational to the understanding of marine ecosystems within the context of climate change.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.