2025-isceplus

Course materials for the 2025 Technical Short Course: InSAR Processing and Analysis (ISCE+) plus reusable local InSAR workflows built around ISCE3, COMPASS, and Dolphin.

Start Here (WSL2 + Linux)

Why: this repo is built around Linux-native geospatial tooling (isce3, compass, dolphin, GDAL).

1. Open a WSL2 Linux shell (Ubuntu) and work inside Linux paths.
2. Keep this repo under /home/... (recommended), not on C: / /mnt/c/....
3. Open VS Code in WSL mode (Remote - WSL) if you use VS Code.

Environment Setup

Why: all scripts assume the isce3-feb conda environment.

From repo root:

cd /home/niels/insar/git/2025-isceplus
mamba env update -n isce3-feb -f envs/isce3-feb.yml
# If mamba is not installed, use:
# conda env update -n isce3-feb -f envs/isce3-feb.yml
conda activate isce3-feb
python scripts/00_patch_rasterio_float16.py

Why this patch step exists:

• Current isce3 builds in this env pin GDAL 3.11.x.
• Some rasterio 1.4.x builds with GDAL 3.11.x miss Float16 dtype mapping (KeyError: 15).
• The patch backports the missing dtype map so Dolphin outputs can be read directly.

Quick checks:

python -V
which python
python -m pip -V

Important:

• Do not run project scripts with /bin/python3.
• Use either conda activate isce3-feb + python ... or mamba run -n isce3-feb python ....

Credentials

Why: ASF (Earthdata) and OpenTopography calls need credentials before search/download/DEM steps.

bash scripts/00_check_credentials.sh

If checks fail, follow:

• knowledge_library/01_credentials_earthdata_opentopography.md

Reusable Project Scaffold

Why: this repo now supports creating new AOI projects without editing Miami files.

Create a new project from template:

python scripts/01_create_project_from_example.py \
  --repo-root . \
  --project-name my_city

For decomposition workflows, create ASC+DSC stacks in one pass:

python scripts/01_create_project_from_example.py \
  --repo-root . \
  --project-name my_city \
  --dual-track

Then follow:

• example_project/README.md (full end-to-end instructions)
• projects/my_city/.../config/processing_configuration.toml (project-specific settings)

Useful setup helpers:

• scripts/02_discover_s1_candidates.py: discover candidate direction/orbit/frame geometry coverage before locking search settings.
• scripts/04_suggest_reference_date.py: suggest a baseline-aware reference date from your searched stack dates.
• scripts/90_decompose_los_velocity.py: decompose ASC/DSC LOS velocity rasters into East/Up velocity rasters.
• scripts/10_plot_ifg_network_qc.py: create interferogram-network QC PNG/JSON from prepared Dolphin inputs.

Config note:

• If your repo has multiple stack configs, always pass --config <.../processing_configuration.toml> explicitly.

Repo Layout

• course/: course notebooks and lesson material (reference/tutorial content)
• envs/: conda environment files
• miami/: current Miami project + shared pipeline scripts
• example_project/: reusable project template
• projects/: generated/custom project workspaces
• knowledge_library/: step-by-step runbooks for recurring tasks
• scripts/: project bootstrap/discovery/credential helpers

Next Docs

• Pipeline run order: miami/README.md
• Generic project workflow: example_project/README.md
• Project-specific status/details: miami/insar/us_isleofnormandy_s1_asc_t48/README.md
• Learning runbooks: knowledge_library/README.md