/eel-pdst-analysis

Analysis of pDST data from European eels

Primary LanguageR

eel-pdst-analysis

Analysis of pDST data from European eels

Pop-off data storage tags (pDSTs) log data (e.g. water temperature and pressure) during a fish's trajectory. After a preprogrammed time, the tag detaches, surfaces, drifts ashore and can be found by, for instance, beachcombers and fishermen who send the tag back to us. Consequently, each datafile contains information related to the trajectory of 1 fish, logging multiple variables (e.g. water temperature and pressure) which can be at different rates. This data allows us to reconstruct a fish's trajectory in the marine environment.

Project structure

R Markdown

R Markdown files and forthcoming HTML files with code and output.

Data

  • /raw: This folder contains the raw download files with temperature and depth data from the archival tags (pDST and PSAT). The folder non-processed-filescontains raw files that have not been processed and put in the interim folder.

  • /interim:

    • /daylogs: The daylog-files contain daily summaries of temperature and pressure data (min, mean and max).
    • /sensorlogs: The sensor-files contain the raw temperature and pressure data. The data from Nieuwpoort (Belgium) are publically available at Zenodo: https://doi.org/10.5281/zenodo.8398240
    • /geolocation_input_files:
      • /input_tagID: The 'tag_ID' corresponds to the actual tag ID of a tag. Each folder contains 4 input files for the geolocation model:
        • tagID_LONG.csv: linear interpolation of longitude between start and end date
        • tagID_PRES.csv: depth data (corrected for depth drift)
        • tagID_TEMP.csv: temperature data
        • tagID_TEMP_F.csv: average temperature at sea surface (first 20 m) and daily max depth
    • /dvm_noon: Files with the calculated noon, based on the DVM behaviour
    • batch_processed_eels_5min.csv: file with all eels for analysis aggregated over 1 min period and with corrected pressure values.
    • data_circadian_5min.csv: batch_processed_eels.csv with circadian phases.
    • data_circadian_tidal_5min.csv: data_circadian.csv with tidal data.
    • data_circadian_tidal_moon_sun_5min.csv: data_circadian_tidal.csv with moon illumination fraction, sun altitude and sun azimuth data.
    • data_depth_diff.csv: batch processed file for which the depth difference is calculated between minima and maxima.
    • data_hourly_migrations.csv: batch processed file with the number of vertical migrations per hour together with hourly summarised environmental data

  • /external:

    • /geolocation_input_files: input files for the geolocation model per eel with lat and lon of release and retrieval or pop-off location
    • /trajectory_data: output from geolocation model with calculated daily lat and lon, and daily summary on temperature and depth (files obtained by David Righton)
    • /dvm_noon: Equation of time applied to files with the calculated noon, based on the DVM behaviour, to calculate coordinates
    • /tidal_old_files: .dat files with incorrect 2018 tidal data. Note that the part in 2019 of 3 eels is correct (obtained by Lianne Harrison (CEFAS))
    • /tidal_2018: .dat files with correct 2018 (tidal obtained by Lianne Harrison (CEFAS))
    • /tidal_data_lianne_harrison: .dat files obtained by Lianne Harrison (CEFAS) containing information on tidal direction and current strength
    • parameters.csv: file containing the parameters to batch process the sensor files
      • start_datetime: moment eel was released (GMT)
      • end_datetime: end point of data of interest; in this case when eel is on the continental shelf (GMT)
        • till DVM
        • till one hour before predation
        • 15 minutes before popoff time
      • bank_datetime: moment eel was on the river bank, i.e. start_datetime - 15 min (GMT)
      • popoff_datetime: moment the tag reached the surface (GMT)
      • UTC: UTC time correction (in hours) to convert to GMT
      • pressure_correction: indicates if a pressure correction needs to be applied (0 = no, 1 = yes)
      • comment: any comments
    • migration_direction.csv: file with the migration direction of the relevant eels for analysis

  • /output: This folder contains visual outputs from specific scripts.

Scripts

! Important note: the clock settings of the pDSTs are the same as Pieterjan's laptop, hence their time setting is UTC+1 (Brussels time zone)!

  • /src:
  1. pdst_functions.R: Functions to serialize/read and tidy pDST datasets
  2. pdst_raw_to_tidy.R: Convert pDST csv files into tidy pDST csv files and store them in /data/interim/
  3. create_temperature_depth_plots.R: Create plots with the logged temperature and depth over time
  4. calculate_noon: Calculate noon based on DVM behaviour. This noon can be used to calculate lat and lon with the equation of time (in excel)
  5. interpolate_longitude.R: Create input file for model: longitude interpolation
  6. create_t_and_p_files.R: Create input file for model: temperature and pressure files
  7. create_sst_file.R: Create file with daily sea surface temperatures (first 20 m depth) and max depths
  8. read_trajectory_data.R: Batch read and process trajectory data obtained via matlab toolbox which is stored in ./data/external/trajectory_data
  9. analyse_migration_speed.R: Calculate and analyse horizontal migration speeds
  10. batch_all_eels.R: Create sensor log file with all eels for analysis, taking into account time settings (UTC), pressure drift and redundant data (e.g. data when not on the shelf, during DVM and predation). Also link eel metadata and the trajectory data (coordinates) obtained via geolocation model to the sensor file.
    • Generate batch_processed_eels_5min.csv in interim folder
    • Generate batch_processed_eels_1hour.csv in interim folder
  11. link_circadian.R: Link circadian phases to the dataset
    • Generate data_circadian_5min.csv in interim folder
    • Generate data_circadian_1hour.csv in interim folder
  12. link_tidal.R: Link tidal data to the dataset (tidal data obtained via Lianne Harrison (CEFAS, UK))
    • Generate data_circadian_tidal_5min.csv in interim folder
  13. link_moon_sun.R: Link illuminated moon fraction, sun altitude and sun azimuth data to the dataset
    • Generate data_circadian_tidal_moon_sun_5min.csv in interim folder
  14. classify_currents.R: classify currents based on surface current data, i.e. eastward, westward, northward and southward currents
    • Generate data_current_phases.csv in interim folder
  15. calc_depth_diff.R: calculate vertical movement distance per hour
    • Generate data_depth_diff.csv in interim folder
  16. calc_hourly_migrations.R: calculate the number of vertical migrations per hour together with hourly summarised environmental data
    • Generate data_hourly_migrations.csv in interim folder
  17. create_plots.R: Create plots for data exploration
  18. visualise_tracking_data.R: Visualise tracking data as animated gif using moveVis R-package. The output is stored in data\output\.
  19. create_actogram.R: Create an actogram to visualise the activity (i.e. vertical movement behaviour) of the eels
  20. create_periodogram.R: Create periodogram to identify main periods in depth time series of the eels
  21. create_acf.R: Create an ACF plot to visualise the activity (i.e. vertical movement behaviour) of the eels
  22. analyse_depth.R: Analyse distance from seabed according to the circadian phases (day and night), current phases (eastward, westward, northward and southward currents)
  23. analyse_depth_diff.R: Analyse depth difference according to the circadian phases (day and night) and current phases (eastward, westward, northward and southward currents)
  24. speed_current_analysis.R: Analyse maximum daily migration speed relative to maximum daily current strengths