Skip to content
/ seamap Public

Seamap rendering based on open-data and MapLibre

3 stars 1 fork Branches Tags Activity
Star
Notifications You must be signed in to change notification settings

prozessor13/seamap

BranchesTags

Repository files navigation

Seamap.java - Planetiler Profile for Nautical Charts

Screenshot Split View

Overview

This Java file implements a Planetiler profile that converts OpenStreetMap data directly into PMTiles vector tiles for nautical charts. Additionally, in the Demopage, all Bathymetry/Contours/Spotsoundings are rendered ondemand via a patched version of the maplibre-contour plugin.

For offline usage on boats, the SignalK Seamap Plugin provides intelligent map tile caching with multi-tier retrieval: filesystem cache (7-day retention), offline PMTiles sectors (~350km² coverage at zoom 6), and online fallback via HTTP range requests. Supports vector tiles (OSM, Seamap) and raster bathymetry (GEBCO, EMOD) with serverside dynamic contour and bathymetry generation.

Features

1. Seamark Base Objects (seamark:* tags)

  • Processes all seamark types: buoys, beacons, lights, landmarks, harbours, etc.
  • Extracts comprehensive attributes:
    • osm_id, type, name, reference, function, category, shape
    • color, color_pattern (with semicolon → underscore conversion)
    • light (abbreviated format), light_color, light_sequence
    • topmark_color, topmark_shape (sanitized)
  • Creates separate layers:
    • seamark_point - Point features
    • seamark_linestring - Line features
    • seamark_polygon - Polygon features with additional label points (PointOnSurface)

2. Derived Seamark Features

Maps standard OSM tags to seamark types:

OSM Tags Seamark Type Category Geometries
route=ferry ferry_route - linestring
waterway:sign=anchor anchorage - point/line/polygon
power=cable + location=underwater cable_submarine power linestring
man_made=pipeline + location=underwater pipeline_submarine substance linestring
man_made=pier mooring pier point/line/polygon
leisure=marina harbour marina line/polygon
leisure=swimming_area|nature_reserve restricted_area leisure value line/polygon
man_made=tower|windmill|gasometer landmark man_made point
man_made=lighthouse lighthouse - point (with full light attributes)

3. Land and Water Layer

Land Layer

  • Downloads and processes global land polygons from osmdata.openstreetmap.de
  • Source: land-polygons-split-4326.zip (~600 MB)
  • Projection: WGS84 (EPSG:4326)
  • Automatic download on first run if not present
  • Creates land layer with polygon features
  • Attributes: None (simple land mask geometry)
  • Buffer: 4 pixels for smooth rendering at tile boundaries

Water Layer

  • Extracts water bodies from OSM (natural=water)
  • Intelligent bathymetry handling: Water polygons are classified based on whether they have real bathymetric survey data (using GEBCO TID grid)
  • Water WITH bathymetry (e.g., IJsselmeer, Bodensee, coastal lagoons):
    • Cut out from land polygons via tile post-processing
    • Allows bathymetry rendering beneath the water surface
    • Not included in final water layer
  • Water WITHOUT bathymetry (e.g., small inland lakes, ponds):
    • Kept in water layer for normal blue water rendering
    • Prevents green interpolation artifacts from missing bathymetry data
    • Attributes: name, water (type: lake, reservoir, pond, etc.)

This approach ensures that:

  • Large navigable waters with hydrographic surveys show detailed bathymetry
  • Small inland waters without surveys render as standard blue water features
  • No confusing green "depth 0" areas appear where no real bathymetry data exists

4. Light Sector Layer

Generates geometric representations of light sectors for navigational lights:

Light Geometries

  • Arcs: Colored sector arcs showing the visible range of each light color
  • Rays: Lines at sector boundaries marking the transitions between colors

Light Types

Type Arc Radius Ray Radius
light_major 0.7 NM (~1296m) 2.5 NM (~4630m)
light_minor 0.4 NM (~741m) 1.2 NM (~2222m)

Attributes

  • osm_id (integer) - Reference to parent seamark
  • type (string) - Parent seamark type (e.g., light_major, light_minor)
  • subtype (string) - Geometry type: arc or ray
  • color (string) - Light color for arcs (e.g., red, green, white)
  • range (string) - Nominal range in nautical miles
  • sector_start (integer) - Start bearing in degrees (0-360, for arcs only)
  • sector_end (integer) - End bearing in degrees (0-360, for arcs only)

Source Tags

Processes seamark:light:N:* tags where N is the sector number:

  • seamark:light:N:colour - Light color
  • seamark:light:N:range - Nominal range
  • seamark:light:N:sector_start - Sector start bearing
  • seamark:light:N:sector_end - Sector end bearing

Example

For a lighthouse with red (0-120°) and green (120-240°) sectors:

  • 2 arc features (one red, one green)
  • 4 ray features (at 0°, 120°, 120°, 240°)

5. Places Layer

  • Extracts natural=bay features
  • Creates point geometry:
    • Points: direct geometry
    • Lines: centroid point
    • Polygons: point on surface
  • Attributes: osm_id, type, subtype, name, reference

6. Default Values for IALA Maritime Buoyage System

Implements automatic defaults for standardized seamark types:

Cardinal Marks

  • North: black_yellow horizontal, 2 cones up, black topmark
  • East: black_yellow_black horizontal, 2 cones base together, black topmark
  • South: yellow_black horizontal, 2 cones down, black topmark
  • West: yellow_black_yellow horizontal, 2 cones point together, black topmark

Isolated Danger

  • Color: red_black_red horizontal
  • Topmark: 2 spheres, black

Safe Water

  • Color: red_white vertical
  • Topmark: sphere, red

Special Purpose

  • Color: yellow

Generic Defaults

  • All beacons: shape = buoyant (transformed from pile)
  • All buoys: shape = pillar
  • Colors with underscore: color_pattern = horizontal

5. Light Abbreviation

Replicates SQL seamark_light_abbr() function:

  • Format: <character>(<group>).<COLORS>.<period>s<height>m<range>M
  • Supports single and multi-light configurations
  • Extracts max range per color
  • Example: Fl(3).WRG.10s15m12M

Usage

Required Dependencies

Download the required JAR files into a jar/ directory:

mkdir -p jar
cd jar

# Planetiler (main dependency)
wget https://github.com/onthegomap/planetiler/releases/download/v0.9.3/planetiler.jar

# TwelveMonkeys ImageIO WebP support
wget https://repo1.maven.org/maven2/com/twelvemonkeys/imageio/imageio-webp/3.10.1/imageio-webp-3.10.1.jar -O imageio-webp.jar
wget https://repo1.maven.org/maven2/com/twelvemonkeys/imageio/imageio-core/3.10.1/imageio-core-3.10.1.jar -O imageio-core.jar

# TwelveMonkeys Common utilities
wget https://repo1.maven.org/maven2/com/twelvemonkeys/common/common-io/3.10.1/common-io-3.10.1.jar -O common-io.jar
wget https://repo1.maven.org/maven2/com/twelvemonkeys/common/common-lang/3.10.1/common-lang-3.10.1.jar -O common-lang.jar

cd ..

Run

Basic usage (downloads OSM data from Geofabrik):

java -Xmx12g -cp .:./jar/planetiler.jar:./jar/imageio-webp.jar:./jar/imageio-core.jar:./jar/common-lang.jar:./jar/common-io.jar Seamap.java \
  --area=croatia \
  --force

With depth data for automatic depth calculation on rocks and wrecks:

java -Xmx12g -cp .:./jar/planetiler.jar:./jar/imageio-webp.jar:./jar/imageio-core.jar:./jar/common-lang.jar:./jar/common-io.jar Seamap.java \
  --area=croatia \
  --depth=data/depth.pmtiles \
  --force

The --depth parameter accepts a PMTiles file with bathymetry data (Terrarium-encoded). Download GEBCO bathymetry:

mkdir -p data
wget https://fsn1.your-objectstorage.com/mtk-seamap/gebco.pmtiles -O data/depth.pmtiles

Memory optimization for large areas (e.g., Germany):

java -Xmx12g -cp .:./jar/planetiler.jar:./jar/imageio-webp.jar:./jar/imageio-core.jar:./jar/common-lang.jar:./jar/common-io.jar Seamap.java \
  --area=germany \
  --storage=mmap \
  --nodemap-type=array \
  --force

Output Layers

Point/Line/Polygon features for all seamark objects

Attributes:

  • osm_id (integer)
  • type (string) - seamark type
  • name (string)
  • reference (string)
  • function (string)
  • category (string)
  • shape (string)
  • color (string)
  • color_pattern (string)
  • light (string) - abbreviated light characteristics
  • light_color (string)
  • light_sequence (string)
  • topmark_color (string)
  • topmark_shape (string)

Note: Polygon features also generate a corresponding point feature for labeling (using point on surface).

Build PMTiles file for the whole planet

This command takes about 1h on a strong hetzner machine (Ryzen 9 & 128 GB RAM)

java -Xmx20g   -XX:+UseParallelGC   -XX:ParallelGCThreads=4   -cp "/app/classes:/app/libs/*:/app/resources:."   Seamap.java   --download   --osm-url=https://planet.openstreetmap.org/pbf/planet-latest.osm.pbf   --maxzoom=14  --output=/data/seamap.pmtiles   --tmp=/data/tmp   --nodemap-type=array   --storage=mmap   --threads=20   --download-threads=10   --http-retries=3   --force

Demo Page (seamap.html)

The seamap.html file provides a complete MapLibre GL JS demonstration of the nautical chart visualization.

Data Sources for the Demopage

1. Base Map

  • Provider: VersaTiles
  • Style: Colorful style with desaturated colors (-30% saturation)
  • Purpose: Base map with streets, labels, and general cartography

2. Seamark Vector Tiles

  • Source: MapToolkit Data Connector
  • Tile Endpoint: https://dataconnector.maptoolkit.net/seamap/seamap/{z}/{x}/{y}.pbf?api_key=seamap
  • Pmtiles Source: https://fsn1.your-objectstorage.com/mtk-seamap/seamap.pmtiles
  • Format: Vector tiles (PMTiles served via API)
  • Max Zoom: 14
  • Layers:
    • seamark - All nautical features (buoys, beacons, lights, etc.)
    • land - Land polygons for coastline
    • light - Light sector geometries (arcs and rays)

3. Seamark Sprites

  • Source: MapToolkit Icons
  • Endpoint: http://icons.maptoolkit.net/seamap
  • Format: Sprite sheets with S-57 nautical symbols
  • Purpose: Icon rendering for buoys, beacons, landmarks, etc.

4. Land Hillshading & Contours

  • Provider: Mapterhorn
  • Tile Endpoint: https://tiles.mapterhorn.com/{z}/{x}/{y}.webp
  • Encoding: Terrarium (RGB-encoded elevation)
  • Max Zoom: 12
  • Features:
    • Hillshading with 0.2 exaggeration
    • Dynamically generated contour lines
    • Contour intervals: 200m/1000m (z7), 100m/500m (z8-9), 50m/200m (z10), 20m/100m (z11), 10m/50m (z12)

5. Bathymetry (EMODnet)

  • Source: MapToolkit Data Connector (EMODnet Bathymetry 2024)
  • Tile Endpoint: https://dataconnector.maptoolkit.net/seamap/emod/{z}/{x}/{y}.webp?api_key=seamap
  • Gebco Pmtiles Source: https://fsn1.your-objectstorage.com/mtk-seamap/emod.pmtiles
  • Emodnet Pmtiles Source: https://fsn1.your-objectstorage.com/mtk-seamap/gebco.pmtiles
  • Encoding: Terrarium (RGB-encoded depth values)
  • Max Zoom: 11 (9 for Gebco)
  • Coverage: European waters (worldwide for Gebco)
  • Features:
    • Bathymetric hillshading (0.2 exaggeration)
    • Dynamically generated depth contours
    • Contour lines at: 0m, 2m, 5m, 10m, 20m, 50m, 100m, 250m, 500m, 1000m, 2000m, 3000m, 4000m, 5000m
    • Depth area fills: 0-2m, 2-5m, 5-10m, 10-20m, 20-50m
    • Spot soundings (depth labels) with 32-pixel grid spacing
    • All Contour/Bathymetry/Spotsounding features a created ondemand in the browser via the maplibre-contours plugin

Technology Stack

  • MapLibre GL JS: Open-source map rendering engine
  • maplibre-contour (v0.2.8): Plugin for dynamic contour generation from DEM tiles
  • VersaTiles Style: Base map styling library
  • Maptoolkit Servers: Serve Seamap and bathymetry tiles

Features Rendered

The demo visualizes:

  • Bathymetry: Colored depth zones with contours and soundings
  • Land: Coastline with terrain hillshading and elevation contours
  • Seamarks: Buoys, beacons, lights with IALA colors and symbols
  • Light Sectors: Colored arcs and rays showing navigational light coverage
  • Navigation: Traffic separation schemes, fairways, anchorages
  • Hazards: Rocks, wrecks with appropriate symbols
  • Infrastructure: Cables, pipelines, platforms, landmarks

Generate Bathymetry Water ID List (Optional)

To enable intelligent water/bathymetry handling, you need to generate a list of OSM water IDs that have real bathymetric survey data. This is done by analyzing the GEBCO TID (Type Identifier) grid, which indicates data source types.

Prerequisites

Install Python dependencies:

pip install geopandas rasterio shapely fiona

Step 1: Download GEBCO TID Grid

# Download GEBCO 2024 TID grid (~90 MB compressed, 4 GB uncompressed)
wget https://www.bodc.ac.uk/data/open_download/gebco/gebco_2024_tid/zip/ -O gebco_2024_tid.zip
unzip gebco_2024_tid.zip

# Convert NetCDF to tiled GeoTIFF (CRITICAL for performance)
gdal_translate -of GTiff \
  -co TILED=YES -co BLOCKXSIZE=512 -co BLOCKYSIZE=512 \
  -co COMPRESS=LZW \
  NETCDF:"GEBCO_2024_TID.nc":tid \
  gebco_tid_tiled.tif

Step 2: Extract OSM Water Polygons

osmium tags-filter planet-latest.osm.pbf \
  wr/natural=water wr/water wr/landuse=reservoir \
  -o water_filtered.osm.pbf

ogr2ogr -f GPKG osm_water.gpkg water_filtered.osm.pbf \
  -sql "SELECT osm_id FROM multipolygons"

Step 3: Run Bathymetry Tagging Script

python3 tag_water_bathymetry.py \
  --water osm_water.gpkg \
  --tid gebco_tid_tiled.tif \
  --output water_with_bathymetry.txt \
  --workers 8

Parameters:

  • --threshold 0.2 - Minimum ratio (20%) of real bathymetry data required (default)
  • --min-area 0.00005 - Minimum polygon area in degrees² (~500m², filters small features)
  • --workers 8 - Number of parallel processing threads

GEBCO TID Values:

  • 0 = Land/interpolated (no real data)
  • 10 = Predicted (satellite altimetry)
  • 11-15 = Direct measurements (multibeam, singlebeam, etc.)
  • 40-44 = Mixed sources
  • 70-74 = Auto-generated / crowd-sourced
  • Any value != 0 indicates some kind of bathymetric data source exists

Output:

  • water_with_bathymetry.txt - OSM IDs (one per line) for water features with real bathymetry data
  • Place this file in the same directory as Seamap.java before running the tile generation

Expected Results

Water bodies that SHOULD have bathymetry (will be cut from land):

  • IJsselmeer, Markermeer (Netherlands)
  • Loch Ness, Lake Geneva (major lakes)
  • Lagune von Venedig (coastal lagoons)
  • Large navigable reservoirs and inland seas

Water bodies that SHOULD NOT have bathymetry (will render as blue water):

  • Small unnamed lakes in forests
  • Ornamental ponds and water features
  • Polder lakes without surveys (Netherlands)
  • Random small reservoirs

Maintenance

Regenerate water_with_bathymetry.txt when:

  • Updating to a new GEBCO release (annually, typically July)
  • Processing a new OSM extract or planet file

Create EMODnet Bathymetry

Download data

Download EMODnet bathymetry dataset (NetCDF format with elevation data for European waters)

wget https://erddap.emodnet.eu/erddap/files/bathymetry_dtm_2024/EMODnet_bathymetry_2024.nc

Download OSM land polygons (global coastline shapefile in WGS84)

wget https://osmdata.openstreetmap.de/download/land-polygons-complete-4326.zip
unzip land-polygons-complete-4326.zip

Process bathymetry data

  1. Extract elevation layer from NetCDF and convert to GeoTIFF
    • Assigns WGS84 coordinate system (EPSG:4326)
    • Converts from NetCDF format to standard GeoTIFF raster
gdal_translate -a_srs EPSG:4326 NETCDF:"EMODnet_bathymetry_2024.nc":elevation bathymetry_elevation.tif
  1. Rasterize land polygons to mask land areas
    • burn 0: Set land pixels to elevation 0 (sea level)
    • init nan: Initialize all pixels as NaN (no data)
    • tr: Target resolution ~0.00104° (~115m at equator, matching EMODnet resolution)
    • te: Extent covering Europe (-36°W to 43°E, 15°N to 90°N)
    • ot Float32: Output as 32-bit floating point for precise elevation values
gdal_rasterize -burn 0 -init nan -a_nodata nan -tr 0.001041666666666 0.001041666666666 -te -36 15 43 89.999999999933507 -ot Float32 -of GTiff -l land_polygons land-polygons-complete-4326/land_polygons.shp land0.tif
  1. Merge land mask with bathymetry data
    • Overlays land0.tif (land=0) onto bathymetry_elevation.tif (ocean depths)
    • Preserves NaN values where no data exists in either source
    • Result: Combined raster with land at 0m and bathymetry for sea areas
gdalwarp -overwrite -ot Float32 -srcnodata "nan" -dstnodata "nan" -of GTiff land0.tif bathymetry_elevation.tif merged.tif
  1. Fill small gaps in data using interpolation
    • md 5: Maximum distance of 5 pixels to search for valid values
    • Interpolates missing data (NaN) using surrounding valid pixels
    • Creates seamless bathymetry raster without gaps
gdal_fillnodata.py -md 5 merged.tif filled.tif
  1. Create raster-dem tiles

This command takes about 10h on a strong Hetzner machine

rio rgbify -v -e terrarium --min-z 3 --max-z 11 -r -2 -j 4 --format webp filled.tif emod.mbtiles
pmtiles convert emod.mbtiles emod.pmtiles

About

Seamap rendering based on open-data and MapLibre

Resources

Readme
Activity

Stars

3 stars

Watchers

2 watching

Forks

1 fork
Report repository

Releases

No releases published

Packages

No packages published

Contributors 2

  •  
  •  

Languages