SINDBAD

Welcome to the repository for Strategies to INtegrate Data and BiogeochemicAl moDels (SINDBAD).

Overview

SINDBAD is a model data integration framework that encompasses the biogeochemical cycles of water and carbon, allowing for extensive and flexible integration of parsimonious models with a diverse set of observational data streams.

The framework provides a unified system for:

• Model Structure: Adaptable ecosystem process representations
• Input Data: Flexible data ingestion and processing
• Observation Data: Versatile observational constraint integration
• Integration Methods: Customizable optimization and assimilation approaches

Repository Structure

This repository contains the Sindbad package sources. The Sindbad package depends on several other registered packages (some maintained in separate repositories), which are installed automatically by Julia’s package manager.

• src/: Sindbad user-facing modules:
  • DataLoaders/: Data loading and preprocessing
  • Setup/: Experiment configuration and setup
  • Simulation/: Terrestrial ecosystem model execution
  • ParameterOptimization/: Parameter calibration and optimization
  • MachineLearning/: ML-assisted surrogates and emulators
  • Visualization/: Plotting and visualization tools
  • Experiment/: High-level experiment orchestration
  • Types/: Type definitions for SINDBAD structures
• ext/: Extension packages for optional dependencies
• examples/: Example experiments and configurations
• docs/: Documentation source files

Internal packages (SINDBAD core):

• SindbadTEM: Core terrestrial ecosystem models, types, and utilities

Related packages (SINDBAD ecosystem):

• ErrorMetrics.jl: Model–observation metrics (maintained in a separate repository)
• TimeSamplers.jl: Time aggregation / sampling utilities (maintained in a separate repository)
• OmniTools.jl: Shared utility toolkit (maintained in a separate repository)

Installation

From a Julia package registry (recommended)

using Pkg
Pkg.add("Sindbad")

From source (development / monorepo checkout)

If you are working from a checkout of this repository:

using Pkg
Pkg.activate("path/to/SINDBAD")
Pkg.instantiate()
using Sindbad

If you are using SSH-based package sources for SINDBAD ecosystem packages, you may need to tell Julia to use the system git client:

export JULIA_PKG_USE_CLI_GIT=true

If you want to develop Sindbad from a separate environment, you can also use:

using Pkg
Pkg.develop(path="path/to/SINDBAD")

Optional dependencies (extensions)

Some functionality is enabled via Julia package extensions (see Project.toml [weakdeps] + [extensions] and ext/):

• NLsolve: enables parts of the spinup workflow (SindbadNLsolveExt)
• Optimization: enables SciML Optimization-based optimizers (SindbadOptimizationExt)
• CMAEvolutionStrategy: enables CMA-ES optimizer bridge (SindbadCMAEvolutionStrategyExt)

To enable an extension, add the corresponding package in the same environment where you use Sindbad:

using Pkg
Pkg.add("NLsolve")  # or "Optimization", "CMAEvolutionStrategy"

For development setup and usage instructions, see CONTRIBUTING.md.

Quick Start

using Sindbad

# Run a forward simulation for an experiment configuration
out = Sindbad.Experiment.runExperimentForward("experiment_config.json")

# `out` contains `out.info`, `out.forcing`, and `out.output`

Documentation

Comprehensive documentation is available at:

• Stable: landecosystems.github.io/Sindbad.jl/stable
• Development: landecosystems.github.io/Sindbad.jl/dev

SINDBAD Contributors

SINDBAD is developed at the Department of Biogeochemical Integration of the Max Planck Institute for Biogeochemistry in Jena, Germany with active contributions from Sujan Koirala, Xu Shan, Jialiang Zhou, Lazaro Alonso, Fabian Gans, Felix Cremer, Nuno Carvalhais.

For a full list of current and previous contributors, see sindbad-mdi.org team page

Contributing

We welcome contributions! Please see CONTRIBUTING.md for guidelines on:

• Development setup
• Code style and conventions
• Submitting pull requests
• Reporting issues

Copyright and License

SINDBAD: Strategies to Integrate Data and Biogeochemical Models

Copyright © 2025
Max-Planck-Gesellschaft zur Förderung der Wissenschaften

For copyright details, see the NOTICE file.

---

License

SINDBAD is free and open-source software, licensed under the European Union Public License v1.2 (EUPL).

---

Your Rights

You are free to:

• Copy, modify, and redistribute the code
• Use the software as a package in your own projects, regardless of their license or copyright status
• Apply the software in both commercial and non-commercial contexts

---

Your Responsibilities

If you modify the code — excluding changes made solely for interoperability — you must redistribute the modified version under the EUPL v1.2 or a compatible license. This ensures the long-term sustainability of the project and supports an open, inclusive, and collaborative community.

---

Disclaimer

This software is provided in the hope that it will be useful, but without any warranty — including, without limitation, the implied warranties of merchantability or fitness for a particular purpose.