BathyMap

BathyMap publishes raw bathymetry survey data as an interactive map on GitHub Pages: https://wrhgroup.github.io/BathyMap/

The raw files live in data/, and the website is built from them on your own computer with one command (BathyMap build), which writes it to docs/. GitHub Pages serves docs/ straight from the main branch, so publishing an update means committing and pushing. There is no GitHub Actions workflow and no Jekyll theme.

What visitors can do

• See where every survey is: labelled location markers when zoomed out, survey tracks closer in, and individual soundings from zoom level 15.
• Choose the basemap: OpenStreetMap (the default), satellite, street, topographic, light grey, dark grey or ocean. Place names and roads can be laid over any of them.
• Hover over or tap a sounding to read its depth, date, time (to a hundredth of a second, as logged) and position.
• Colour the soundings by water depth (the default) or by another column: bed elevation, GNSS height, time of day, distance along line, platform speed, roll, pitch or RTK correction age. The colour scale is fitted to the soundings in view or fixed for all data.
• Browse the lines in a list grouped by location and survey day. The list opens collapsed, can be filtered by location or by text, and has checkboxes that show or hide lines on the map one at a time or all at once.
• Open a line to see its statistics, platform, sensor and other metadata, and download it as the original .dep file or as a CSV with decimal-degree coordinates.
• Open any survey day in 3D: the measured soundings with an interpolated bed between them, to rotate and zoom, with an adjustable vertical exaggeration, and the interpolated bed as a CSV download.
• Draw a box or a polygon on the map and download every sounding inside it as a CSV.
• Download every line in one CSV, and share a link straight to a line (the address ends with #<line id>).
• Return to the full view with the Zoom to all data button under the zoom buttons.

Repository layout

BathyMap/
├── 1_site_config.json      settings for the website (see "Configuration")
├── data/                   the raw survey files, organised in survey folders (see "Adding new data")
│   ├── metadata.json
│   └── 2026-10-16/
│       ├── L0001_2026-10-16-00-24-42.dep
│       └── L0002_2026-10-16-00-52-33.dep
├── docs/                   the published website, written by "BathyMap build"; never edit it by hand
├── site/                   the website template (index.html, app.js, style.css, and in site/vendor/ the Leaflet
│                           map library, the three.js 3D library and the fonts), copied into docs/ on every build
├── BathyMap/               the Python package behind the BathyMap command
│   ├── cli.py
│   └── site_builder.py
├── tests/                  pytest tests and small sample files
└── setup.py

Setting up your computer (once)

You need Python 3.9 or newer and Git. A separate conda environment keeps BathyMap away from the packages in the Anaconda base environment, some of which were compiled for an older NumPy and print warnings. In the Anaconda Prompt, go to the repository folder (the one that contains setup.py) and run:

conda create -n bathymap python=3.12 -y
conda activate bathymap
pip install -e .

This installs the BathyMap command with its requirements: click, pandas, numpy, scipy (for the interpolated bed) and pyproj (for converting coordinates to metres). After updating BathyMap to a version with new requirements, run pip install -e . again in the environment. The -e keeps the command linked to this folder, so changes to the code take effect without reinstalling. In every new terminal, run conda activate bathymap before using the command.

Publishing on GitHub Pages (once)

The repository is wrhgroup/BathyMap. In it, open Settings → Pages. Under Build and deployment, set Source to Deploy from a branch, choose branch main and folder /docs, and save. GitHub publishes the site a minute or two later, and again after every push that changes docs/.

To add a "Data repository" link to the website, set repository_url in 1_site_config.json to https://github.com/wrhgroup/BathyMap.

Adding new data

1. Copy the raw files into a new survey folder in data/. To name the location, or if the platform or sensor differs, add a metadata.json file (see below).

2. From the repository folder, build the website:

   BathyMap build
   

   The command reads 1_site_config.json in the current folder. (BathyMap build -f other_config.json uses another settings file.) It prints one block per data file: its location, platform, sensor, number of soundings and depth range, and how many soundings were removed by cleaning. It ends with "website written to ...".

3. Optionally, preview the site (see "Previewing").

4. Commit and push the data and the rebuilt website:

   git add data docs
   git commit -m "Add survey 2026-11-03"
   git push
   

If new files are in the repository but not on the website, the build was skipped or docs/ was not committed. Run the build, then commit and push again. After a push, a hard refresh (Ctrl+F5) makes sure the browser loads the new version.

Organising the data folder

Keep one folder per survey (usually one per survey day), named with the date first so folders sort in time order, and drop the raw files in unchanged:

data/
├── metadata.json                         applies to all data (platform, sensor)
├── 2026-10-16/
│   ├── metadata.json                     applies to this survey only, e.g. its location
│   ├── L0001_2026-10-16-00-24-42.dep
│   └── L0002_2026-10-16-00-52-33.dep
└── 2026-11-03_credit-river/
    ├── metadata.json
    ├── L0001_2026-11-03-14-02-10.dep
    ├── L0002_2026-11-03-14-40-55.dep
    └── L0002_2026-11-03-14-40-55.json    applies to this one file only

• Use lowercase letters, digits and hyphens in folder names, without spaces. Any depth of sub-folders works.
• Don't rename or move a file once it is published. The folder and file name form the line's identifier, which is used in shared links (for example …/BathyMap/#2026-10-16__L0001_2026-10-16-00-24-42) and in the download file names.
• The list on the website is grouped by the location in the metadata and by the survey date and area name recorded inside each file, not by folder.
• The line name shown on the website is the part of the file name before the first underscore (L0001).

Metadata: location, platform, sensor and anything else

Each line is described by entries from JSON files, merged from the general to the specific. A later level overrides only the entries it sets and inherits the rest:

1. survey_metadata in 1_site_config.json (empty by default).
2. metadata.json files, from the top of data/ down to the file's own folder. Each applies to its folder and every folder inside it.
3. A .json file with the same name as the data file, next to it. It applies to that file only.

The repository's data/metadata.json describes the current platform and sensor:

{
    "platform": "CHCNAV Apache 4",
    "platform_type": "Uncrewed surface vessel (USV)",
    "sensor": "Apache 4 single-beam echosounder"
}

A survey folder's metadata.json usually only names the location:

{
    "location": "Credit River"
}

• location groups the lines in the list, fills the location menu, labels the markers on the zoomed-out map and appears in the line details. Spell it the same way every time you return to a place, because lines are grouped by the exact text. Lines without one are listed under "No location given".
• platform, platform_type and sensor have their own rows in the line details. Platform and sensor are also written as columns in every CSV download. If no platform is given at any level, the antenna type recorded in the file header is used.
• Any other entry (for example operator, project, weather or notes) appears in the line details and can be found with the text filter.

Previewing

BathyMap build
python -m http.server -d docs 8000

Then open http://localhost:8000. Stop the server with Ctrl+C before the next build, because while it runs Windows (and OneDrive) can keep files in docs/ locked. Opening docs/index.html directly from disk does not work, because browsers block the data files from loading that way.

Using the map

Disclaimer

Each time someone opens the site, a disclaimer appears over it and stays until they click I understand (or press Esc). It says the data are provided as they are, without warranty, may contain errors and must not be used for navigation, and that no liability is accepted for their use. Read the disclaimer in "About this data" shows it again. Its title, paragraphs and button text are set by disclaimer in 1_site_config.json, so they can be changed without touching the code; show_disclaimer: false stops it from appearing on opening. The wording follows the disclaimers of other bathymetry and open-data publishers (GEBCO, USGS, Ontario Open Data), but it is not legal advice; have it checked by your institution if needed.

Basemaps

The layers button at the top right of the map chooses the basemap and the overlays. None of them needs an account or an API key:

Basemap	default_basemap	Source	Closest native zoom
OpenStreetMap (default)	openstreetmap	OpenStreetMap standard tiles	19
Satellite	satellite	Esri World Imagery	19
Street map	street	Esri World Street Map	19
Topographic	topographic	Esri World Topographic Map	19
Light grey	light	Esri Light Gray Canvas	16
Dark grey	dark	Esri Dark Gray Canvas	16
Ocean	ocean	Esri World Ocean Base	16

The Place names and Roads overlays (Esri reference layers) can be ticked over any basemap. Beyond a basemap's closest native zoom, its last tiles are enlarged, so the light grey, dark grey and ocean maps turn soft when zoomed in to the soundings. The visitor's choice of basemap and overlays is remembered in their browser, so the default only applies on a first visit.

OpenStreetMap's tiles are drawn by volunteer-run servers, and close-up tiles of rarely viewed places can take a moment to appear the first time. CARTO basemaps have required an API key since 25 September 2026. CARTO offers free keys for non-commercial use at https://carto.com, and with one in carto_api_key, CARTO Voyager and CARTO Positron are added to the menu. (The key is visible in the published website, as with any web-map key.) Free tile services can change their terms; if one stops working, the others are unaffected.

Finding the data

Any line drawn smaller than 40 pixels on screen is shown as a labelled marker with its location, number of lines and number of soundings. Markers closer than 70 pixels merge into one labelled "N locations". Clicking a marker zooms to it. The map opens showing all the data, and Zoom to all data returns to that view.

The line list

The list opens with every location collapsed (or every survey day, when no locations are named). The arrows fold and unfold each location and survey day, and Collapse all / Expand all do it for the whole list. Clicking the name of a location or survey day zooms the map to its lines without opening it; only the arrow opens and closes it. The checkboxes show or hide lines on the map: All lines at the top (only the lines matching the current filter, when one is set), one per location, one per survey day and one per line. A checkbox with a dash means some of its lines are hidden. Selecting a line opens its details above the list, and the list keeps its place.

3D view of the bed

Each location and each survey day has a 3D button in the list, and each line's details have 3D view of this survey day. A location's 3D view interpolates all its survey days together, for one bed of the whole location; the panel says when it combines several dates, because the bed may have changed between them. A location surveyed on a single day shares the files of that day. The 3D view replaces the map. The Map and 3D bed tabs at the top switch between the two. 3D bed opens the survey in the map view: of the lines shown inside the view, the location with the most lines (or, for lines without a location, the survey day with the most), framed on the area of the map view when that is smaller than the survey. With no survey in view, a note says so. Map, choosing a line, clicking the name of a location or survey day, and Zoom to all data return to the map. Unchecking lines that the open 3D view is built from also returns to the map, zoomed to that area, with those lines hidden.

The view shows two point clouds, coloured by water depth:

• Measured soundings (points with a dark rim): the soundings of the survey day, thinned like the map (max_map_points_per_line).

• Interpolated bed (plain points): the bed estimated between the soundings. The build triangulates the soundings of the survey day (or location) and interpolates the depth linearly onto a regular grid (bed_grid_spacing, 0.5 m by default), in metres in the local UTM zone. A triangle of the triangulation is filled when it joins two different passes of the survey and none of its sides is longer than bed_max_gap (30 m by default). The bed is therefore continuous between neighbouring passes, but it does not bridge wider gaps, and it is never extended beyond the outermost passes, because the triangles only exist between soundings. Nothing is filled beside a lone pass, which shows only its measured soundings: values there could only be carried sideways from that one pass. Setting bed_max_distance above 0 adds a strip of that width around every sounding; beside a lone pass, its values come from triangles made of two neighbouring soundings of the pass and one sounding of the nearest other line, however far away.

  Two soundings count as different passes when they come from different lines or survey days, or when the boat travelled more than bed_pass_ratio (2 by default) times the straight distance between them along its track, as between the legs of a back-and-forth survey. A bend within one pass does not count, so the inside of a curve sailed once is not filled.

Drag to rotate, scroll or pinch to zoom, and right-drag or two-finger-drag to move. A basemap is laid as a flat sheet at the water surface around the survey, with the surveyed area cut out of it so the bed shows through. The surveyed area is the outline of the survey: every triangle joining soundings with no side longer than bed_max_gap, including the holes inside the bed where nothing is interpolated, so no part of the basemap is left inside it. Water surface draws a transparent surface over the same outline, so it never extends beyond the survey. The Basemap at the water surface menu offers the same basemaps as the map, and None (plain background) for the bed alone over a grid. Until the visitor picks one there, it follows the basemap of the map; their choice is then remembered in their browser. The tiles are downloaded by the visitor's browser when the 3D view opens, like the map's, and are credited in the corner.

Survey areas are much wider than they are deep, so the depths are exaggerated. The 3D view opens with the exaggeration at which the deepest point lies a tenth of the survey's width below the water surface (for example 9× for the Credit River survey, about 160 m across and 1.8 m deep), between 1× and 30×, or with the number set in vertical_exaggeration; the slider changes it between 1× and 60×. The panel at the bottom right sets the vertical exaggeration, the point size, which clouds are shown, and a transparent water surface. Oblique, From above and From the side reset the camera. Hovering over a point says whether it is a measured sounding (with its line) or an interpolated estimate, and gives its depth and, for an estimate, its distance from the nearest sounding. With Fade estimates far from soundings (on by default), interpolated points fade towards the background with distance from the nearest sounding, reaching their faintest at a third of bed_max_gap, so the parts of the bed that rest on few measurements look less certain. The colour scale follows the This view / All data choice of the map: this survey day's depths, or the scale shared by all surveys.

Interpolated bed (CSV) downloads the grid as Latitude, Longitude, Interpolated depth and Distance to nearest sounding (metres). The interpolated values are estimates, not measurements: a single-beam survey measures one depth per sounding, and the bed between survey lines is only as good as the line spacing allows. The 3D view and the CSV say so. Each survey day is interpolated on its own, so surveys of the same place on different dates are never blended. The 3D library is only downloaded when a visitor opens a 3D view.

Colouring the soundings

The Colour by menu in the legend switches between the map variables set in map_variables (see "Configuration"). With the default settings and the CHC files, the menu offers:

Group	Variable	From the column	Use
Bathymetry	Water depth	Depth	The depth as logged by the echosounder (the draft recorded in the file header is listed in each line's details).
	Bed elevation	Height − Depth	The shape of the bed (see "Heights are ellipsoidal").
	GNSS height	Height	The height of the GNSS antenna or water surface (see "Heights are ellipsoidal").
Survey	Time of day	Time	The order in which the soundings were collected.
	Distance along line	Trail Distance	How far the platform had travelled since the first sounding of the line. The raw column counts from the start of the survey session in the CHC software and carries on from one file to the next (L0001 starts at 80 m, L0002 at 1,206 m), so the first value of each line is subtracted. The CSV keeps the raw values.
	Platform speed	Speed	Fast runs give sparser soundings.
	Flow velocity	Flow Velocity	Shown once a survey records it (zero in the current data).
Quality	Roll, Pitch	Roll, Pitch	Tilt of the platform, converted from radians to degrees. A tilted single-beam transducer measures off vertical. Blue is negative, red positive, white zero.
	Correction age	Diff Delay	Age of the RTK corrections in seconds; older corrections mean less accurate positions.
	Horizontal and vertical precision	HRMS, VRMS	Shown once a survey records them (zero in the current data).
	GPS state	GPS State	Shown once it varies (every sounding is RTK fixed in the current data).

The remaining columns are not offered. Azimuth and Trail Direction are directions, which wrap around at north and suit no colour scale, and Line Distance, Base ID and Data ID are not measurements of the survey. Every column is still in the CSV downloads.

The two buttons under the colour scale set its range:

• This view (the default) stretches the colours over the 2nd to 98th percentile of the soundings in view, updated after each pan or zoom, so small differences stand out wherever you look.
• All data keeps one scale for every survey, so colours can be compared between places.

Exporting an area

The two dashed buttons below Zoom to all data select soundings:

• Box: press, drag and release over the soundings.
• Polygon: click or tap each corner, then click the first corner, double-click, or press Finish.

The bar at the top of the map then shows how many soundings were selected, from how many lines, their depth range and time span, with a Download CSV button. Esc cancels a shape being drawn. The export is cut from the full CSV of each line, so it contains every selected sounding even where the map shows a thinned line. Its columns are Line, Datetime, Latitude, Longitude, Depth, Platform, Sensor, followed by every other column of the original file. Hidden lines are left out.

Configuration (1_site_config.json)

Paths are relative to the JSON file, which stays at the top of the repository. Every key is optional except data_directory. A key that is left out takes its default, and the build prints a line saying so.

Key	Default	What it does
data_directory	required	Folder with the raw files.
site_template_directory	site	Website template copied into the output.
output_directory	docs	Where the website is written; must match the Pages folder. It is emptied on each build, and the build refuses to empty a folder that a previous build did not create (it looks for the .bathymap-build file it leaves there).
site_title, site_description	BathyMap, empty	The title is shown in the header and the browser tab, the description in "About this data".
organisation, contact, licence	empty	Shown in "About this data" when set.
show_disclaimer	true	Shows the disclaimer each time the site is opened.
disclaimer	no warranty, not for navigation, no liability	The disclaimer: title, paragraphs (a list, one text per paragraph) and button.
repository_url	empty	GitHub address of the repository. Adds a "Data repository" link, and is needed when copy_raw_files is false.
repository_branch	main	Branch used for the raw-file links when copy_raw_files is false.
file_extensions	[".dep"]	File types read from the data folder.
accepted_gps_states	[] (keep all)	For example [4] keeps RTK-fixed soundings only.
max_map_points_per_line	3000	Soundings drawn on the map per line. Downloads always contain every sounding.
max_track_points_per_line	200	Points in the simplified track drawn when zoomed out.
copy_raw_files	true	Copies the raw files into docs/ so they download from the website. With false, the download links point to the files in data/ on GitHub instead, which keeps the published site smaller.
write_combined_csv	true	Writes data/all_lines.csv with every line, for the "Download all lines" button.
survey_metadata	{}	Metadata applied to every line before the metadata.json files are read.
metadata_file_name	metadata.json	Name of the per-folder metadata files.
time_reference	as logged by the instrument	Describes the clock of the times, for example UTC.
depth_reference	as logged by the instrument	Describes what the depths are measured from.
height_reference	as logged by the instrument	Describes the vertical reference of the GNSS heights; shown as "Heights" in "About this data".
depth_colour_range	automatic	[min, max] in metres for the "All data" depth colour scale; automatic uses the 2nd to 98th percentile of all depths.
map_variables	the 13 variables listed under "Colouring the soundings"	The "Colour by" menu (see below).
default_basemap	openstreetmap	Basemap on a visitor's first visit: openstreetmap, satellite, street, topographic, light, dark or ocean (or carto-voyager and carto-positron with a CARTO key).
carto_api_key	empty	CARTO key; when set, the CARTO basemaps are added to the menu.
start_with_list_collapsed	true	Opens the line list collapsed.
interpolate_bed	true	Interpolates the bed of each survey day for the 3D view. With false, no 3D view is offered.
bed_grid_spacing	0.5	Grid spacing of the interpolated bed, in metres.
bed_max_distance	0	When above 0, the bed is also filled within this distance (metres) of every sounding, even beside a lone pass, where the values are carried sideways from that pass.
bed_max_gap	30	Largest gap (metres) between passes that the bed is filled across. Set it a little above the usual line spacing of the surveys.
bed_pass_ratio	2	Two soundings of one line count as different passes when the track between them is more than this many times their straight distance apart.
vertical_exaggeration	auto	Starting vertical exaggeration of the 3D view: auto (the deepest point a tenth of the survey's width below the water surface, 1× to 30×) or a number such as 5.
bed_max_points	250000	Largest number of interpolated points per survey day. When a survey would need more, the build widens the grid spacing and says so.

map_variables

Each entry describes one item of the "Colour by" menu. A few of the defaults:

"map_variables": [
    {"column": "Depth", "label": "Water depth", "unit": "m", "colours": "depth", "decimals": 3, "group": "Bathymetry"},
    {"column": "Height - Depth", "label": "Bed elevation", "note": "GNSS height minus water depth. Above the ellipsoid, not sea level", "unit": "m", "colours": "elevation", "decimals": 3, "group": "Bathymetry"},
    {"column": "Time of day", "label": "Time of day", "unit": "", "colours": "sequential", "decimals": 0, "format": "time", "group": "Survey"},
    {"column": "Roll", "label": "Roll", "note": "Logged in radians, shown in degrees", "unit": "°", "scale": 57.29578, "colours": "diverging", "decimals": 1, "group": "Quality"}
]

• column is a column of the raw file, the difference of two columns written as "A - B", or "Time of day" (the time of each sounding).
• label, unit, decimals and note (optional, shown under the scale) control how it is shown. group places it under a heading in the menu. format: "time" shows the values as clock times.
• scale (optional) multiplies the values, for example 57.29578 to turn radians into degrees.
• from_line_start: true (optional) subtracts the value of each line's first sounding, so the values start at zero on every line.
• colours is depth (pale shallow to dark deep), elevation (the same ramp reversed, so a low bed is dark), sequential (purple to yellow, kept distinct so it is never read as depth) or diverging (blue below zero, white at zero, red above, with zero kept in the middle of the scale).
• range (optional, for example [0, 3]) fixes the "All data" scale; otherwise it spans the 2nd to 98th percentile of all values.
• Water depth is always the first entry. A column that no file has, or that has the same value everywhere (for example GPS State when every sounding is RTK fixed), is left out of the menu, and the build log says so.
• Derived and converted values (bed elevation, time of day, roll and pitch in degrees) are computed for the map only; the CSV downloads keep the original columns.
• Points files grow with each variable (about 210 KB per line of 2,000 soundings with the defaults). They are only loaded when a line is selected or zoomed in to.

Heights are ellipsoidal, not above sea level

The Height column of the CHC files is the height measured by the GNSS receiver. No geoid model was applied in the survey software, so it is the height above the ellipsoid, the smooth mathematical shape of the Earth that GNSS works with, and not the height above sea level used on maps. The difference between the two, the geoid height, depends on the location: in Canada it is tens of metres, and it changes by several metres across a province. GNSS height and bed elevation are therefore far from the contours of the topographic basemap, and the gap is not the same at every survey site. The official geoid height of any point in Canada is given by NRCan's GPS-H tool (https://webapp.csrs-scrs.nrcan-rncan.gc.ca/geod/tools-outils/gpsh.php?locale=en).

The relative values are unaffected. In the 2026-10-16 survey Height stays within 20 cm (standard deviation 3 cm), so Height − Depth shows the shape of the bed correctly. It may also include the height of the GNSS antenna above the water (the file header records 0.20 m as "Antenna H"). To obtain heights above sea level (CGVD2013 in Canada), set the survey project's height system to CGVD2013 with the CGG2013a geoid model in the CHC software before exporting.

The website says this in three places: the "Heights" line in "About this data" (from height_reference), and the notes under the colour scale for GNSS height and bed elevation.

What the build does

1. Reads 1_site_config.json.
2. Finds every file with a listed extension in data/ and its sub-folders.
3. Empties docs/ and copies the template from site/ into it.
4. For each file: reads the header (area, date, antenna type, draft, sound velocity) and the soundings; converts the coordinates from degrees-minutes-seconds to decimal degrees; gives every sounding a full date and time (handling lines that cross midnight); removes invalid soundings; and merges the metadata.
5. Writes, for each line, the map points (docs/data/points/), a full-resolution CSV (docs/data/csv/), a copy of the raw file (docs/data/raw/) and its rows in docs/data/all_lines.csv.
6. Interpolates the bed of each survey day (date, area and location) and of each named location (all its survey days together), and writes it to docs/data/bed/: a JSON file for the 3D view and a CSV file for download.
7. Writes docs/data/index.json, the summary the website reads first: lines, survey days with a 3D view, statistics, simplified tracks, metadata, colour scales and download links.

What is cleaned

Soundings are removed when the depth is missing or not positive, a coordinate is missing or out of range, or (when accepted_gps_states is set) the GPS state is not accepted. The build log reports how many soundings were removed from each file. The raw file download is always the untouched original. GPS states follow the NMEA convention used by the receiver: 1 = GPS, 2 = DGPS, 4 = RTK fixed, 5 = RTK float.

Tests

From the repository folder, with the environment active:

pip install pytest
pytest

The tests build a small site from the sample files in tests/files/ into tests/files/_test_site/, which Git ignores.

Limits

GitHub recommends keeping a repository under 1 GB, a published Pages site cannot exceed 1 GB, and Git rejects single files over 100 MB. Git stores identical files only once, so the raw copies in docs/data/raw/ add nothing to the size of the repository. The CSV and map files in docs/ do add roughly two to three times the size of the raw data, and each survey day adds its interpolated bed (about 1.3 MB for the 2026-10-16 survey at 0.5 m; a coarser bed_grid_spacing shrinks it). When the archive grows large, set copy_raw_files to false to shrink the published site. For a much larger archive, keep the raw files in GitHub Releases or a DOI archive such as Zenodo and keep this repository for the map.

Third-party components

The map uses Leaflet 1.9.4 (BSD-2-Clause), the 3D view uses three.js r147 (MIT), and the site uses the fonts Barlow Semi Condensed, Source Sans 3 and JetBrains Mono (SIL Open Font License). All of them are included in site/vendor/ with their licences, so the website loads nothing from font or script CDNs. Basemap tiles come from OpenStreetMap and Esri, and are credited in the corner of the map.