bpnsdata 0.2.0

A package to add marine environmental data to a geopandas dataframe

BPNSdata

bpnsdata is a package to add environmental data to a geopandas DataFrame. There is no support for multiindex columns, so one level has to be selected or dropped out before using it. Right now only Belgian Part of the North Sea data is available for all the classes. However, some classes are not restricted to the bpns and can be used to add environmental data to other parts of the world.

Install

pip install bpnsdata

Installing the requirements can be a bit tricky, so it might fail during installation if you are working on Windows. If that is the case,we recommend to install FIRST the following packages (in this order) by downloading the wheels of:

• GDAL
• rasterio
• Fiona

You can follow this tutorial if you're not familiar with wheels and/or pip: https://geoffboeing.com/2014/09/using-geopandas-windows/

Environmental data

Environmental data can be added by specifying it in the env_vars variable when calling the main class SeaDataManager. To do so, it is necessary to have gps information, which can be stored in a .gpx file, in the "waypoints" or "track_points" layer. It can also be loaded as a csv or a shp file. Then the algorithm finds the point which is closest in time for row of the dataframe. The available data sources are:

• csv: Static geo data in a csv file. Files to be provided by the user
• time: Information about the moment of the day and the moon
  • moon cycle
  • day moment
• emodnet: wcs data from EMODnet
  • shipping density
  • bathymetry
• raster: raster data (tiff images)
  • seabed habitats
  • habitat suitability
• griddap: RBINS data from the erddap server
  • sea surface
  • wave information
• wrakken_bank: shipwreck information
• meetnet_vlaamse_banken: read weather data from the buoys of the meetnet vlaamse banken
• ais: AIS data from the AIS hub from VLIZ. Only access when connected to the VPN of VLIZ for the moment.

For easier running of the classes, there is a main class called SeaDataManager, which allows to run all the desired environmental variables in one line of code.

EMODnet

Entry point to download map data from EMODnet using WCS. Coverage to be checked in EMODnet, but larger than BPNS. The implemented classes so far are:

Shipping (class ShippingData)

Shipping activity from https://www.emodnet-humanactivities.eu/ Adds the route density or the shipping intensity from the month of the deployment to the dataset, considering the location, the year and the month. It adds the columns:

• route_density
• ship_density (depending on the layer type selected)

Bathymetry (class BathymetryData)

Adds the mean bathymetry (https://www.emodnet-bathymetry.eu/) layer considering location (no time considered) The output column is:

• bathymetry

Raster Data

Raster Data represents geographical data. Only BPNS available The two outputs are:

Seabed habitats (class SeabedHabitatsData)

Adds the sea habitat (https://www.emodnet-seabedhabitats.eu/). The output columns are:

• seabed_habitat
• substrate

Benthic habitats (class BenthicHabitatsData)

Habitat suitability map from the publication ([1]V. Van Lancker, G. Moerkerke, I. Du Four, E. Verfaillie, M. Rabaut, and S. Degraer, “Fine-scale Geomorphological Mapping of Sandbank Environments for the Prediction of Macrobenthic Occurences, Belgian Part of the North Sea,” Seafloor Geomorphology as Benthic Habitat, pp. 251–260, 2012, doi: 10.1016/B978-0-12-385140-6.00014-1.). The closest point from the maps is added to the each point of the dataset. The output column is:

• benthic_habitat

ERDDAP RBINS Data

Sea State Data from RBINS (https://erddap.naturalsciences.be/erddap/index.html). Coverage to be checked in the RBINS erddap website, but restricted to North Sea. In this version only the tables BCZ_HydroState_V1 and WAM_ECMWF are implemented.

Sea Surface (class SeaSurfaceData)

The data is added from the table: BCZ_HydroState_V1.

• surface_baroclinic_eastward_sea_water_velocity
• surface_baroclinic_northward_sea_water_velocity
• sea_surface_height_above_sea_level
• sea_surface_salinity
• sea_surface_temperature
• surface_baroclinic_sea_water_velocity

Sea Bottom (class SeaBottomData)

The data is added from the table: BCZ_HydroState_V1.

• bottom_baroclinic_eastward_sea_water_velocity
• bottom_baroclinic_northward_sea_water_velocity
• bottom_upward_sea_water_velocity

Wave Data (class WaveData)

The data is added from the table: WAM_ECMWF Output columns:

• hs: wave height in cm
• tm_1: wave period

Time Data (class TimeData)

Data Related to time series. It adds the time of the day (day, night, twilight dawn...) and the moon phase. The calculation is done using skyfield (https://rhodesmill.org/skyfield/). Coverage in all the world. The output columns are:

• moment_day (twilight, dawn, day, night)
• moon_phase (in radians)

Csv Data (class CSVData)

Static data that is stored in a csv, with a lat and a lon columns (names to be given). It returns the closest point of all the csv, the distance to it, the coordinates and also other columns selected by the user with the specified suffix.

Wrakken Bank (class WrakkenBankData)

Will add information about the closest shipwreck. The data is extracted from https://wrakkendatabank.afdelingkust.be/. Following information will be added:

• shipwreck_distance: Distance to closest shipwreck
• shipwreck_lat
• shipwreck_lon
• shipwreck_name

Meetnet Vlaamse Banken

Read the available weather forecast at the closest buoy from https://api.meetnetvlaamsebanken.be/V2-help/. Attention! To be able to use this feature you need to have a user registered at Meet Net Vlaamse Banken. You can do it for free from their webpage. Then you need the username and the password. You can pass it directly to the created objects, but if you want to use them in the SeaDataManager you will have to add the username and the password as environmental variables (username_bank and password_bank). So far, rainfall (NSI) and average wind speed at 10 m (WVC) are implemented. It adds to the DataFrame a column with the value of the data, the id of the specified buoy and the distance to the buoy. The id of the buoy is represented by the sum of the location id + the data id. i.e., in the buoy OMP, the id for precipitation is OMP+NSI=OMPNSI

Rain (class RainData)

Rainfall in NSI at the closest buoy

Wind (class WindData)

Average wind speed at 10 m from the surface, at the closest buoy

AIS (class AISData)

AIS data from the AIS hub from VLIZ. Only access when connected to the VPN of VLIZ for the moment. Adds the columns:

• ais_total_seconds: total cumulative seconds when a ship was there
• ais_n_ships: total number of ships
• ais_total_seconds_distance_weighted: total cumulative number of seconds, weighted according to the distance of each ship

Usage

Possible ways of loading the data. By default, all the classes read the column 'datetime' as the column from the GeopandasDataFrame where the time information is stored, but the user can select another column by specifying the datetime_column argument.

import bpnsdata 
import pandas as pd 
import numpy as np

# When the desired df is already on a gpx or a csv with coordinates (in this case, imagine the gpx itself 
# contains the rows to analyze
geofile = 'data/VG.gpx'
geodf = bpnsdata.SurveyLocation(geofile).geotrackpoints

# Could also be done directly using geopandas: 
geodf = geopandas.read_file(geofile)

# Could be that we have a df (here a random one) and we want to add a geolocation to it
# Create a random dataframe to work with
time_index = pd.date_range(start='2020-10-12 11:35', end='2020-10-12 12:00', freq='m', tz='UTC')
random_data = np.random.randint(5, 30, size=10)
df = pd.DataFrame(random_data)
df['datetime'] = time_index

All the classes can be used separately, by calling each class in its own. First declare the class with the desired parameters, then call the object with the df as an argument.

import bpnsdata

# For example, for shipping: 
shipping = bpnsdata.ShippingData(layer_name='routedensity', boat_type='all')
df_env = shipping(geodf)

Use of the SeaDataManager

The SeaDataManager can be used when multiple env parameters have to be added. Then the user needs to list which ones need to be added. These names are the names of the available classes but with all small letters and an underscore separating the words, and removing the Data at the end.

For example:

• TimeData -> time
• HabitatSuitabilityData -> habitat_suitability
• SeaSurfaceData -> sea_surface

import bpnsdata

# Define the seadatamanager
env_vars = ['shipping', 'time', 'wrakken_bank', 'habitat_suitability', 'bathymetry'
            'seabed_habitat', 'sea_surface', 'sea_bottom', 'sea_wave', 'rain', 'wind', 'ais', 'sea_wave_north_sea']
manager = bpnsdata.SeaDataManager(env_vars)

# If the data is without geometry, then:
geodf = manager.add_geodata(df, gpx_file)

# Once the data has geometry:
df_env = manager(geodf)

If specific parameters want to be passed, then you can pass a dictionary when calling the SeaDataManager, where the key has to be the name of one of the env_vars and value is a dictionary with key, value as the parameters that can be passed to the call function of that class.

import bpnsdata

# Define the seadatamanager
env_vars = ['shipping', 'time', 'wrakken_bank', 'habitat_suitability', 'bathymetry'
            'seabed_habitat', 'sea_surface', 'sea_bottom', sea_wave', 'rain', 'wind', 'ais']
manager = bpnsdata.SeaDataManager(env_vars, {'ais': {'buffer': 20000}})

# Call the seadatamanager 
env_df = manager(geodf)