Background & Summary

In the frame of fisheries assessment, echosounder data are routinely collected around the world1,2, and acoustic data are being used increasingly to study several different features of aquatic ecosystems3,4,5,6. However, there is still a disconnect between biological oceanography and fisheries research, and often cruises that sample the deep ocean (i.e., the mesopelagic zone from 200 to 1,000 m and beyond) do not collect echosounder data even if the instruments are available onboard.

The 2010 Malaspina circumnavigation expedition aimed at studying the biodiversity and the impact of global change on the deep ocean. The interdisciplinary project collected samples and data across multiple disciplines - from physics and chemistry to genomics and biodiversity7 - producing a number of new insights about the ocean8,9,10,11,12,13. Echosounder data collected during the Malaspina expedition have resulted in numerous publications related to mesopelagic fish biomass and behavior8,14,15,16. However, the data still contain a huge amount of information across a range of spatial (meters to oceans basin) and temporal (seconds to seasonal trends) scales that could be exploited in different ways: the objective of this paper is to provide adequate access to those data.

Methods

Figure 1 presents the track of the eight-month cruise, and Table 1 provides the detail of the legs and dates. On a routine basis R/V Hesperides sailed at an average speed of 11 knots from around 3 pm to 4 am (local time). The vessel arrived on station at around 4 am daily to carry out sampling operations at a fixed point for about 11 hours.

Fig. 1
figure 1

Cruise track and integrated backscatter at different stations (NASC, daytime 200 to 1000 m).

Full size image
Table 1 Dates and starting points of the 7 legs of the Malaspina cruise.
Full size table

Acoustic measurements were carried out continuously using a Simrad EK60 echosounder), operating at 38 and 120 kHz (7° beamwidth transducers) with a ping rate of 0.5 Hz. Unfortunately, the 120 kHz failed during the first leg of the cruise and only 38 kHz data were collected. Echosounder observations were recorded down to 1000 m depth. The echosounder files are in the proprietary Simrad raw format and can be read by various softwares (e.g., LSSS, Echoview, Sonar5, MATECHO, ESP3, echopype, pyEcholab). GPS locations and calibration constants are imbedded in each file.

Additionally, daytime data integrated over 2 m vertical bins from 200 to 1000 m depth are provided as Nautical Area Scattering Coefficient (NASC). Each “voxel” is the average of all cleaned and validated data recorded over that depth range, in a time period starting 8 hours before the start of the station (defined as start of the CTD cast) and ending 8 hours after the start of the station, with only data recorded in the period between 1 hour after local sunrise and 1 hour prior to local sunset accepted (i.e., during local daytime hours, but removing crepuscular periods when vertical migration of biota is strong). The relatively long interval over which data were accepted around each station was chosen since the station sampling resulted in noisy acoustic data,, a long interval was therefore chosen to ensure valid data on all stations.

Finally, summaries of per station daytime and nighttime acoustic data (omitting data recorded within 1 hour of sunrise and sunset) are provided. The data fields in this file are station date, latitude and longitude, and per day and night average NASC 200–1000 m, average NASC 0–1000, weighted mean depth (WMD) of NASC 200–1000 m, migration amplitude, NASC day-to-night ratio and migration ratio.

Data Records

The raw data are available at PANGAEA (https://doi.org/10.1594/PANGAEA.92176017). The daytime data integrated over 2 m vertical bins are available at (https://doi.org/10.1594/PANGAEA.92308718) and the summary file is available at (https://doi.org/10.1594/PANGAEA.92661919).

Technical Validation

The echosounder was calibrated before departure (30th of November 2010, close to Mazarron port, water temperature ca. 17 °C) and values of the peak transducer gain (G0) and Simrad correction factor (Sa) for both frequencies were updated in the ER60 software following the standard target method20. Briefly, a 38.1 mm diameter tungsten carbide sphere was positioned under the vessel at a range of 12 m using nylon cable, and moved systematically through the acoustic beams of both transducers (38 and 120 kHz). The ER60 software compares measured values of the sphere’s target strength at each position in the beam with model predicted values, and estimates the transducers’ calibration parameters (G0, Sa and beam widths) for the given operational settings (see Table 2). Data are stored in raw proprietary Simrad format (.raw,.idx,.bot files).

Table 2 Calibration settings and results.
Full size table

Usage Notes

The main objective of this paper is to allow researchers full access to the data so they can analyze them with objectives that might be completely different from the ones they were collected for. Therefore, the raw data are provided so that each researcher may analyze them using the appropriate methodologies that will differ from those used in previous studies. However, we recommend paying attention to some of the specificities of the cruise that are important when filtering the data. In particular and depending on the objectives data should be filtered for noise following available techniques21,22,23. During the Malaspina expedition, on-station sampling took place from early morning to the beginning of the afternoon, whereas underway navigation occurred from the afternoon and through the night. Therefore, most of the daytime echo sounder data correspond to a stationary situation, and the level of noise is different from day to night (see materials and methods in Irigoien et al., 2014). The data were subjected to different types of noise along the survey. The most common noise was caused by the propeller and/or wave-hull collisions occurring at vessel speeds over 11 knots and were seen as long spikes on the echogram (example in Figs. 2 and 3, see also supplementary figure 7 in Irigoien et al.8 for filtering results). It is recommended to treat these data, especially for depths deeper than 100–200 m, using appropriate filters24. On some occasions, data collected near the coast, in shallow waters, presented interference from other acoustic equipment onboard. This was seen as typical short “flecks” present in pings separated by constant intervals, gradually increasing its range of appearance on the echogram. This type of noise can also be removed by means of impulsive noise filters24. Some parts of the data were also affected by bad weather conditions, this typically seen on the echogram as backscattering losses in some consecutive pings. This can also be treated with well-established filters as the attenuated signal removal. Another issue to be highlighted is that, due to technical problems in some small parts of the tracks, the 38 kHz frequency was in passive mode. Our previous work being focused in the mesopelagic, the 120 kHz data were not used. A preliminary analysis of the 120 kHz data showed to have a high noise to signal ratio. We have included them here as it is not possible to evaluate if useful for specific objectives or using different filtering approaches, however caution is recommended if using the 120 kHz data. Finally, 120 kHz data were not recorded during the first leg due to malfunctions of the transponder.

Fig. 2
figure 2

Example of echogram affected by noise (vertical spikes) for average vessel speed over 11 kn in Leg 2 (18/01/2011 22:06–22:46 h).

Full size image
Fig. 3
figure 3

Example of echogram affected by interference from other acoustic equipment near the coast, in shallow waters, during Leg 4 (11/04/2011 21:26–22:04 h).

Full size image