Data

Saildrone Data Quality Validated by Peer-Reviewed Papers

To establish confidence in Saildrone fit for purpose data, we have sought out and developed relationships with recognized experts in the fields of meteorology and oceanography.

By
Saildrone
,
on
May 6, 2020

Saildrone’s global fleet of unmanned surface vehicles (USVs) are designed to sail the world’s oceans in the most extreme conditions collecting atmospheric and oceanographic data above and below the sea surface. But it’s not enough to simply collect data. To establish confidence in the Saildrone USV platform as a scientific tool, the data collected must be of the highest scientific quality, fit for purpose, with a clear and robust data chain, from sensor calibration to raw observations, processing, and final distribution.

Since the company’s inception, Saildrone has sought out external partnerships and worked closely with established meteorological and oceanographic experts to develop the vehicle sensor suite through rigorous laboratory testing and several years of at-sea comparisons to correlate Saildrone data with established ship and buoy standards (Meinig, Burger, et al., 2019). Initial quality control, drift, calibration, and platform effects are carefully managed to ensure that each sensor and essential variable is properly understood, and calibration and validation are tracked during the hardware and software design cycles.

These public-private partnerships have yielded several peer-reviewed papers describing the Saildrone platform and validating the quality of a specific type of data collected, for example, satellite cal/val, carbon uptake, air-sea heat fluxes, fish biomass, monitoring ocean currents, and meteorological and oceanographic (metocean) measurements.

Saildrone data is available on community portals, thanks to our many partnerships: NOAA ERDAP, NASA PO.DAAC, WMO GTS, EMODNET, COPERNICUS, SOCAT, ICES, ICOS. Raw data is also publicly available for re-analysis via our own data portal, data.saildrone.com (DSC). Data is distributed in NetCDF (Network Common Data Form) format, which include climate and forecast (CF)-compliant metadata—units, calibration dates, the manufacturer and serial number of the instrument and where it was mounted on the vehicle—to describe in detail all procedures undertaken to meet the highest level of scientific standards.

Each of the papers highlighted below represents a partnership between Saildrone and members of the science community. Papers are organized according to their discussed metric and sensor(s). This list will be updated as new papers are published.

Sea surface salinity (SSS) and sea surface temperature (SST)

A team of researchers led by the National Aeronautics and Space Administration (NASA)’s Jet Propulsion Laboratory (JPL) used measurements collected during Saildrone’s 2018 Baja California mission to validate remotely sensed SSS and SST, which measured using the onboard conductivity, temperature, depth (CTD) sensors at .6 meter depth.

Animation comparing Saildrone and satellite SST and SSS data collected during the 2018 Baja California mission.

“Unmanned surface vehicles (USVs), such as Saildrones, provide a novel sampling technique that allows resolving mesoscale and submesoscale processes and, thus, validating the capability of remote sensing data to capture the spatial variability observed in coastal regions.”

Jorge Vazquez-Cuervo, Jose Gomez-Valdes, et al., “Using Saildrones to Validate Satellite-Derived Sea Surface Salinity and Sea Surface Temperature along the California/Baja Coast,” Remote Sensing, 2019.

"Comparisons to satellite data and model analyses highlight the value of Saildrone measurements for future algorithm and numerical model improvements, especially in complex and data sparse oceanic regions."

C. L. Gentemann, Joel P. Scott, et al., "Saildrone: Adaptively Sampling the Marine Environment," Bulletin of the American Meteorological Society, Vol. 101, Issue 6, June 2020

"[Saildrones] can finish observations of atmosphere and ocean variations on mesoscale or smaller scales in fairly short times without disturbing the variations. Therefore, [saildrones] are considered to reveal more exact mesoscale or smaller scale oceanic effects on the atmospheric boundary layer than other methods such as by using only research vessels, satellites, and data assimilation techniques."

A. Nagano, K. Ando, "Saildrone-observed Atmospheric Boundary Layer Response to Winter Mesoscale Warm Spot Along the Kuroshio South of Japan," Prog Earth Planet Sci 7, 43 (2020)

Air-sea fluxes

Two Saildrone USVs participated in the Salinity Processes in the Upper-ocean Regional Study 2 (SPURS-2) field campaign, sailing around the Woods Hole Oceanographic Institution (WHOI) SPURS-2 buoy and with the RV Revelle to validate Saildrone measurements for air-sea fluxes.

“The ship-saildrone-buoy comparison reveals the agreement and instantaneous differences between platforms in relation to separation distance and a passing subsynoptic weather system … This validation suggests that the saildrone can be a valuable autonomous platform for observing air-sea surface heat and momentum fluxes (through bulk algorithm) in the tropical Pacific.”

saildrone buoy comparisons
Saildrones vs. Woods Hole Oceanographic Institution Air-Sea Interaction METeorology System measurements for downwelling shortwave and longwave radiation, air temperature, wind speed and direction, relative humidity (RH), sea surface temperature, and sea surface salinity. Blue lines are SST and SSS measured by the saildrone ASVCO2 SBE CTD. Tair/RH of both buoy and saildrones were adjusted to 2.4 m, wind adjusted to 5 m. Buoy SST/SSS measurements were at 0.95 m, saildrone at 0.6 m. (Zhang et al. 2019)

Dongxiao Zhang, Meghan F. Cronin, et al., “Comparing Air-Sea Flux Measurements from a New Unmanned Surface Vehicle and Proven Platforms During the Spurs-2 Field Campaign,” Oceanography, Vol. 32, No. 2.

“Comparisons between the drone data and surface flux buoys show good agreement, confirming that this platform can make climate-quality meteorological and oceanographic observations.”

Megan F. Cronin, Kathleen A. Donohue, et al., “Abrupt Fronts Embedded in Tropical Instability Waves Observed by Saildrones,” American Geophysical Union, Fall Meeting 2018.

Carbon

Saildrone worked with Australia’s Commonwealth Scientific and Industrial Research Organisation (CSIRO) to deploy two USVs to Maria island off Tasmania to test planning protocols and platform maneuverability, and carbon sensor suite (pCO2, pH, dissolved oxygen) measurement comparisons with a Maria Island National Reference Station moored buoy and the research vessel Investigator.

“At CCS sites, Saildrones are able to measure both surface water properties as well as meteorological data … These data sets are able to provide real-time information on the changing environment, and provide a platform to enable detection and investigation of anomalies.”

Andreas Marouchos, Bronte Tilbrook, et al., “Environmental Assessments of Offshore Carbon Capture and Storage (CCS) Sites Using Unmanned Surface Vehicles (USV),” 14th Greenhouse Gas Control Technologies Conference Melbourne October 21 – 26, 2018 (GHGT-14).

Fish biomass & marine mammals

Saildrone has worked with NOAA on several annual missions to deploy USVs equipped with Simrad WBAT echo sounders in addition to the standard sensor suite to survey the presence of the North Pacific right whale, make acoustic observations of walleye pollock in the Bering Sea, and perform focal follows of northern fur seals. Backscatter observed by the USVs has been compared to backscatter observed by a noise-reduced research vessel during two “follow-the-leader” comparisons.

northern fur seals backscatter bering sea
(A) Total backscatter (sA) recorded by the saildrone during northern fur seal focal follows. Black circles: location of the Saildrone for each dive plot. (B − D) Examples of backscatter with northern fur seal dive behavior overlaid (black line): (B) deep daytime dives, near the sea floor; (C) short, shallow dives occurring at night; (D) mixed  shallow and deep dives at night. Echosounder data are presented from the nearest location (displaced by 5 − 7 km) that matched the time of day the fur seal was diving (day vs. night). In (B) and (C), dive and echosounder data were collected simultaneously. In (D) the closest location with night echosounder data was sampled 24 h prior. (Kuhn et al. 2020)

“Our results show that Saildrones can approach satellite tracked marine predators from long distances and follow them over extended periods while collecting oceanographic and prey data.”

Carey E. Kuhn, Alex De Robertis, et al., “Test of Unmanned Surface Vehicles to Conduct Remote Focal Follow Studies of a Marine Predator,” Marine Ecology Progress Series, Vol. 635: 1–7 (2020).

“Overall, SD detected equivalent backscatter to DY (to within 10%), which is consistent with the agreement in previous comparisons of acoustic measurements from a USV and a research vessel (Swart et al., 2016), different vessels (De Robertis and Handegard, 2013), and different instruments on the same vessel (Macaulay et al., 2018).”

Alex De Robertis, Noah Lawrence-Slavas, et al., “Long-term measurements of fish backscatter from Saildrone unmanned surface vehicles and comparison with observations from a noise-reduced research vessel,” ICES Journal of Marine Science, July 6, 2019.

“Relative to data collected on Oscar Dyson (and other reference platforms), generally the Saildrone data were of high quality and demonstrated the potential use of this platform to augment acoustic fishery surveys.”

Calvin W. Mordy, Edward D. Cokelet, et al., “Advances in Ecosystem Research: Saildrone Surveys of Oceanography, Fish, and Marine Mammals in the Bering Sea,” Oceanography, August 9, 2017.

Saildrone collaborated with NOAA’s Northwest and Southwest Fisheries Science Centers to observe and collect data on zooplankton, fish, marine mammals, and seabirds throughout the California Current Ecosystem during a 100-day mission from Vancouver to San Diego. Five saildrones equipped with Simrad WBT-Mini echo sounders traveled some 18,500 nautical miles, working alongside the NOAA research vessel Reuben Lasker.

“For this survey, however, potential inaccuracies 58 in the apportioning of CPS backscatter to species could amount to only fractions of 1% of the estimated Northern Anchovy and Pacifc Sardine biomasses in the areas sampled by both platforms.”

Kevin L. Stierhoff, Juan P. Zwolinski, and David A. Demer, “Distribution, Biomass, and Demography of Coastal Pelagic Fishes in he California Current Ecosystem During Summer 2018 Based on Acoustic-Trawl Sampling,” U.S. Department of Commerce, NOAA Technical Memorandum NMFS-SWFSC-613

saildrone satellite data comparison
Locations of the Saildrone vehicle ChlF measurements taken off the western coast of Mexico and the U.S. The shading of the vehicle’s trajectory indicates the ChlF measured with the NPQ correction applied, where appropriate. The green and blue markers indicate MODIS-Aqua and VIIRS-SNPP matchups with the saildrone, respectively. (Scott et al., 2020)

“NPQ-corrected and uncorrected Saildrone ChlF measurements are compared against satellite estimates of Chl to evaluate the utility of applying an NPQ-correction to Saildrone in vivo ChlF measurements and to constrain the in situ fluorometer’s calibration via contemporaneously collocated HPLC pigment measurements.”

Joel P. Scott, Scout Crooke, et al., “Correcting non-photochemical quenching of Saildrone chlorophyll-a fluorescence for evaluation of satellite ocean color retrievals,” Optics Express, Vol. 28, Issue 3 (2020).

Oil presence using natural fluorescence

Saildrone worked with the Ecosystem Impacts of Oil and Gas Inputs to the Gulf (ECOGIG) consortium to locate natural oil seeps in the Gulf of Mexico and study how fine-scale environmental variables like wind speed and surface currents affect the size and duration of oil slicks created by natural seeps. A Saildrone USV equipped with a Laser Diagnostics Remote Optical Watcher (ROW) was deployed from Cocodrie, LA. In-situ data collected by the saildrone was compared to Landsat 8 satellite images.

“Projecting the Saildrone track onto the Landsat 8 satellite image indicated that the vehicle had crossed narrow oil slicks. The ROW sensor would therefore alternate between clean surface water and brief exposure to patches of floating oil. In Fig. 8C, the orange signal line exceeding the green lower threshold line indicates fluorescence response from oil films. High-resolution Saildrone observations detected the presence of surface oil with high precision. They also revealed that features observed in the Landsat 8 satellite image were oil slicks and not look-alikes.”

Samira Daneshgar Asl, Dmitry S. Dukhovskoy, et al., “Hindcast modeling of oil slick persistence from natural seeps,” Remote Sensing of Environment, 189 (2017) 96–107.

Metocean

During Saildrone’s first scientific mission, a 97-day cruise to study the effect of sea ice melt on the Bering Sea shelf, SD 128 rendezvoused three times with NOAA RV Oscar Dyson for instrument comparisons, including photosynthetically active radiation (PAR), air and sea temperature, wind speed and direction, humidity, and barometric pressure.

saildrone ship data comparison
Time series of ship-to-saildrone comparisons. Ship values (usually black curves) are plotted every minute, and saildrone values (usually red curves) are plotted every 10 minutes in (a)–(d) and every minute in (e)–(f). (a) PAR, (b) air temperature (ship: black, saildrone: red), water temperature (ship: green, saildrone: blue and cyan), and skin temperature (saildrone: magenta), (c) wind speed, (d) wind direction (from), (e) salinity, and (f) chlorophyll-a concentration. (Cokelet et al., 2015)

“The Saildrones performed reliably in the harsh Bering Sea environment. Comparisons with shipboard measurements showed good agreement, inspiring confidence in these new instrument platforms. Using these unmanned surface vehicles, scientists directed months of sampling in real time to measure the changing ocean conditions of moving, melting sea ice and freshwater pulses from the Yukon River.”

Edward D. Cokelet, Christian Meinig, et al., “The Use of Saildrones to Examine Spring Conditions in the Bering Sea: Instrument Comparisons, Sea Ice Meltwater and Yukon River Plume Studies,” October 2015.

Join the working groups

Saildrone’s commitment to quality control is ongoing. The public involvement of the science community with our growing network of vehicles is a vital part of ensuring that data quality, delivery, and access meet important benchmarks and target key needs of the global ocean observing system (GOOS).

To that end, Saildrone is establishing formal working groups focused on critical areas of ocean observation: surface fluxes, biogeochemistry, fisheries acoustics, and data management. We will continue our close collaboration with members of these working groups (as well as the greater scientific community) to develop and advance pre- and post-mission sensor calibration requirements, CF-compliant data formats, and evolution and evaluation of our data delivery pathway to ensure the highest possible level of data quality.

To join a working group, contact us. Or, read all Saildrone science papers.


Resources:

Christian Meinig, Eugene F. Burger, et al., “Public–Private Partnerships to Advance Regional Ocean-Observing Capabilities: A Saildrone and NOAA-PMEL Case Study and Future Considerations to Expand to Global Scale Observing,”Frontiers in Marine Science 6:448, accessed August 29, 2019