S2GOS-APPS

s2gos-apps is the integration layer of the S2GOS framework. It connects scene generation (s2gos-generator) and observation simulation (s2gos-simulator) into ready-to-run processes that can be executed through a GUI, a CLI, or directly from Python.

Each process is a Python function decorated with @registry.process(). This single definition is enough to expose it as an OGC API endpoint, a GUI form in Jupyter, and a CLI command — powered by the s2gos-controller stack (procodile, wraptile, s2gos-client).

Key components

Module	Role
`registry.py`	`ProcessRegistry` singleton — all processes register here on import
`service.py`	`LocalService` — wraps the registry as a FastAPI / OGC API server
`cli.py`	Generates a CLI (`s2gos_apps`) from the registry using `procodile`
`processes/`	Process definitions grouped by site (Gobabeb, Frascati, …) and common operations
`gen_util.py`	Shared generation logic — runs the `SceneGenerationPipeline` DAG
`sim_util.py`	Shared simulation logic — builds `SimulationConfig`, runs `EradiateBackend`

Available processes

Site-specific configuration processes

Each site provides a pair of processes that expand a few high-level parameters (location, size, observation time) into full configuration files. This is still very early but will be developed into the use cases for teh S2GOS Project

Process ID	Title	Description
`gobabeb-generation-config`	Gobabeb Generation Config	Scene generation config for the Gobabeb PICS site (Namibia)
`gobabeb-simulation-config`	Gobabeb Simulation Config	Simulation config for Gobabeb with configurable GMT hour and samples
`frascati-generation-config`	Frascati Generation Config	Scene generation config for Frascati (Italy) with vegetation placement
`frascati-simulation-config`	Frascati Simulation Config	Simulation config for the Frascati scene
`kairouan-generation-config`	Kairouan Generation Config	Scene generation config for Kairouan (Tunisia) with vegetation
`kairouan-simulation-config`	Kairouan Simulation Config	Simulation config for the Kairouan scene
`pisa-generation-config`	Pisa Generation Config	Scene generation config for San Rossore / Pisa (Italy) with vegetation
`pisa-simulation-config`	Pisa Simulation Config	Simulation config for the Pisa scene
`pnp-generation-config`	PNP Generation Config	Scene generation config for Patagonia National Park (Chile) with heterogeneous atmosphere
`pnp-simulation-config`	PNP Simulation Config	Simulation config for the PNP scene

Common processes

These processes accept a configuration file path and run the full pipeline.

Process ID	Title	Description
`common-generation`	Scene Generation	Runs the generation pipeline from a config JSON — produces meshes, textures, and a scene description YAML
`common-simulation`	Scene Simulation	Runs the Eradiate radiative transfer simulation from a scene description and simulation config

Demo processes

Process ID	Title	Description
`pnp_gui/generation`	Scene Generation Demo	PNP GUI demo with seasonal variations — runs generation pipeline directly
`pnp_gui/simulation`	Simulation Demo	PNP GUI demo with multiple observation types — runs simulation pipeline directly
`upscaling-demo`	Upscaling Demo	Experimental upscaling workflow (not yet implemented)

Installation

You can install the default and development version using:

# default
pixi install
# dev
pixi install -e dev

You can then install the data needed by Eradiate by running.

pixi run apps-init

See the Eradiate documentation on a guide on how to Configure the data install location.

Configuration

The S2GOS package suite can be configured by having a settings file in your current working directory or any parents to it.
The file should be named "s2gos_settings.yaml"
Here is a template example:

# s2gos_settings.yaml
common:
    search_paths:
        - "./resources/data"
        - "./data"

generator:
    dataset:
        dem:
            name: "Copernicus-DEM-30"
            crs: "EPSG:4326"
            type: indexed-geotiff
            root_directory: "/path/to/dem/tiles"
            index_path: "/path/to/dem_index.feather"
            path_column: "path_dem"
            variable_name: elevation

        landcover:
            name: "ESA Worldcover 2021"
            crs: "EPSG:4326"
            type: zarr
            path:
                value: "s3://bucket/worldcover.zarr"
                cid: "my_s3_creds"
            variable_name: landcover

    files:
        material_config: "./materials.json"

See the common and generator configuration pages for additional information.
Secrets can be provided either through environment variables or a .secrets.yaml file. See the credentials page for details on how to set this up.