cross-sensor-cal

Cross-Sensor Calibration turns raw NEON AOP hyperspectral flight lines into analysis-ready reflectance, Parquet tables, and QA artifacts in a single restart-safe workflow. The pipeline covers download, ENVI export, BRDF + topographic correction, cross-sensor convolution, and reporting so teams can focus on analysis instead of data wrangling.

Looking for the old quickstart?

The expanded quickstart now lives at Quickstart with both CLI and Python API walkthroughs.

Get the package

=== "Install from PyPI" bash python -m venv cscal-env source cscal-env/bin/activate # Windows: cscal-env\Scripts\activate pip install --upgrade pip pip install cross-sensor-cal

This installs the CLI entry points (`cscal-download`, `cscal-pipeline`, `cscal-qa`, and friends) plus the Python API. The
wheels are published for Linux and macOS on Python 3.10–3.12; other platforms can build from source.

=== "Install from source" bash git clone https://github.com/earthlab/cross-sensor-cal.git cd cross-sensor-cal python -m venv .venv source .venv/bin/activate pip install -e .[dev]

Source installs expose the latest development features and editable code. See [Environment Setup](env-setup.md) for Conda
instructions and notes about GDAL/PROJ, Ray shared memory, and documentation previews.

Process your first flightline

1. Create a base folder where downloads and derived products will be written: bash BASE=output_demo mkdir -p "$BASE"
2. Run the consolidated pipeline on one or more NEON flight lines. Replace placeholders with the site, month, product, and flight IDs you need: bash cscal-pipeline \ --base-folder "$BASE" \ --site-code NIWO \ --year-month 2023-08 \ --product-code DP1.30006.001 \ --flight-lines \ NEON_D13_NIWO_DP1_L019-1_20230815_directional_reflectance \ NEON_D13_NIWO_DP1_L020-1_20230815_directional_reflectance \ --max-workers 2
3. Inspect results inside $BASE/<flight_stem>/—you should see ENVI exports, corrected cubes, sensor-specific resamples, Parquet tables, and the QA panel (<flight_stem>_qa.png). Re-run the same command at any time; completed stages log ✅ ... (skipping) and are left untouched.
4. Optional: generate fresh QA panels without rerunning the full pipeline. bash cscal-qa --base-folder "$BASE"

Prefer a fully scripted flow? The Quickstart shows the matching Python API example using go_forth_and_multiply().

Troubleshooting & support

• Start with the curated Troubleshooting guide for Ray shared memory issues, filename validation errors, ENVI mismatches, iRODS hiccups, and more.
• Browse the FAQ for configuration and workflow questions like resuming runs or adding new sensors.
• Use --verbose on CLI commands or read the per-flightline logs inside each output directory to pinpoint stage failures.
• When filing issues, include the console output (with the [flight_stem] prefixes) and your configuration overrides to speed up triage.

Explore the documentation

Topic	Start here
Step-by-step walkthrough	Quickstart
Environment and dependencies	Environment Setup
CLI usage and examples	CLI & Examples
Pipeline architecture	Pipeline Stages
Outputs and file formats	Pipeline Outputs and Schemas
Extending sensors & stages	Extending the pipeline
Validation and QA	QA panel, Validation checks

Release highlights

• Per-flightline master table written as <prefix>_merged_pixel_extraction.parquet
• QA panel <prefix>_qa.png is emitted after the merge during full runs

root@d52d22c64421:/workspace/cross-sensor-cal#