Globato

Domo Arigato, Multi-Resolution Globato.

Globato is the user-facing geospatial engine of the Continuous-DEMs ecosystem. It is designed for the rapid development, blending, and processing of high-accuracy Topo-Bathy Digital Elevation Models (DEMs).

Built on top of the fetchez (orchestration) and transformez (horizontal and vertical datums) libraries, globato abstracts away the complexity of DEM development. It allows users to generate massive, seamless DEMs using declarative YAML recipes or intuitive command-line tools—all while scaling infinitely on standard hardware.

Key Features

• Infinite Memory Scaling: globato uses a pure-Python, generator-based streaming architecture to process massive rasters and point clouds chunk-by-chunk, eliminating out-of-memory crashes.
• Declarative Pipelines: Define your data sources, regions, and processing hooks in a simple YAML recipe, ensuring 100% reproducibility.
• API Caching: Remote API queries (NOAA, Copernicus, etc.) are hashed and cached locally. Tweak your blending weights or hillshade parameters and re-run your DEM in seconds without re-downloading data.
• On-The-Fly Datum Shifts: True 3D projection and vertical datum transformations (e.g., MLLW to NAVD88) applied directly to data streams.

Installation

Install via pip (this will automatically pull in the fetchez and transformez dependencies):

pip install globato

The globato CLI

globato provides a powerful, user-friendly Command Line Interface to generate DEMs, run community recipes, manipulate rasters, and more.

1. recipe (The DEM Engine) Generate, execute, and batch-process DEM recipes.

# Build a recipe on the fly from the terminal
globato recipe build -R -120/-119/34/35 -E 1s mbdb copernicus 

# Build a recipe and save it to YAML to learn the syntax
globato recipe build -R loc:"Miami" -E 1/3s mbdb+rq:threshold=50 --save-only

# Run a saved recipe
globato recipe run my_dem_recipe.yaml

2. pointz (The Point Cloud Engine) Stream, filter, and reproject point clouds using standard UNIX pipelines.

# Download multibeam data, filter it, project to Web Mercator, and save to XYZ
globato pointz run mbdb -R loc:"Tampa" -F rq:threshold=10 -T EPSG:3857 > clean.xyz

# Pipe existing local data through Globato's statistical outlier filter
cat raw_survey.xyz | globato pointz run - -F outlierz > clean.xyz

3. viz (The Plotting Engine) Generate instant visualizations for sanity checks without opening QGIS.

# Render a beautiful, georeferenced colored hillshade from a DEM
globato viz hillshade my_dem.tif output_hs.tif --cmap etopo --blend soft_light

# Open an interactive 3D window to inspect a local point cloud
globato viz points raw_lidar.laz --3d

# Fetch live multibeam data, apply a filter, and highlight the rejected noise in red!
globato viz points mbdb -R loc:"San Diego" -F rq:threshold=5 --outliers

4. raster (The Grid Engine) Perform advanced morphological operations and blending on existing grids.

# Blend a high-res coastal DEM into a lower-res background auxiliary grid
globato raster blend hi_res.tif blended_out.tif --aux background.tif --blend-dist 50

5. fetch (The Download Engine) Discover and download raw elevation data directly from federal/global APIs.

globato fetch list --search lidar
globato fetch run tnm -R -120/-119/34/35 -O ./my_data

6. region (The Spatial Engine) Perform spatial math, geocoding, and tileset generation.

# Geocode a city and get the bounding box
globato region echo loc:"Seattle, WA"

# Split a massive bounding box into a 0.25-degree GeoJSON tileset for batch processing
globato region split loc:"California" --size 0.25 -O cali_tiles.geojson

The Vision & Community

The transition to this modern, modular architecture is just the beginning. Our vision is to cultivate a thriving open-source community around high-accuracy elevation modeling.

We welcome contributions! Whether you want to write a new data-fetching module, build a custom visualization hook, or simply share a YAML recipe for your local coastline, check out our GitHub Repository or join our Zulip Chatspace to get involved.