esa-snap-s1tbx-gptSNAP 9–13
ESA SNAP · Graph Processing Tool

Headless SNAP gpt,
installed by conda.

A repackaging of the official ESA SNAP installers that puts the Graph Processing Tool and the Sentinel-1 / SAR stack into a conda package — for reproducible command-line, server, CI, notebook, and pyroSAR workflows. No GUI installer, no Docker, no snappy/jpy.

SNAP9 – 13 OSlinux · win · osx-arm CHANNELsarforge CONDA
Install Full docs → SAR · InSAR · Polarimetry · Earth Observation
Install

One conda command

The package version equals the SNAP version — pick the release you need.

mamba create -n snap13 -c sarforge -c conda-forge esa-snap-s1tbx-gpt=13.0.0
conda activate snap13
gpt -h

conda works too; mamba just solves large environments faster.

Second installs are near-instant — conda caches the package, so a new environment hardlinks it rather than re-downloading ~1 GB of SNAP.
What's inside

The full SAR stack, nothing optical

SAR engine + s1tbx + rstb

All SAR sensors (Sentinel-1, TerraSAR-X, ALOS, RADARSAT, ICEYE, …), calibration, SAR processing, InSAR, polarimetry, classification.

Optical pruned

Sentinel-2/3 (s2tbx/s3tbx) and smostbx removed to keep the package lean.

Conda's own JDK

Linux/macOS use conda-forge openjdk; Windows keeps SNAP's bundled runtime. No separate Java install, no jpy bridge.

pyroSAR-ready

The standard SNAP layout is preserved and the real snap/gpt launchers sit on PATH, so ExamineSnap() auto-detects it.

Published builds

Trusted matrix on sarforge

A storage-conscious set of platforms per SNAP version.

SNAPlinux-64win-64osx-arm64
9.0.0●——
10.0.0●——
11.0.0●——
12.0.0●——
13.0.0●●●
pyroSAR

Zero-config discovery

pyroSAR finds SNAP by locating the snap launcher on PATH — which this package puts there.

from pyroSAR.examine import ExamineSnap

snap = ExamineSnap()
print(snap.gpt)   # resolves inside the conda env

For InSAR phase unwrapping, add snaphu: mamba install -c conda-forge snaphu.

Documentation

Read more

Install & usageCommands, version selection, pyroSAR checks, troubleshooting. Packaging rationaleWhy old SNAP versions matter and why the build matrix is intentional. Jupyter usageDriving gpt and pyroSAR from notebooks. Maintainer guideBuild numbers, local builds, publishing, Pages setup.
Unofficial packaging. Not affiliated with or endorsed by ESA. SNAP is developed by ESA and contributors and licensed under GPL-3.0; the packaging code here is Apache-2.0. See the repository NOTICE.