Scope & Status
Out of scope
Explicitly excluded, so expectations stay defensible in front of a government or donor reviewer:
- Rain attenuation and frequency-band selection — a first-order decision in the tropics, and not resolvable with geometry alone.
- Real anomalous propagation (ducting, measured refractivity gradients) —
only the effective earth-radius factor
kis parameterized. - Sea clutter, ground clutter, RF interference, spectrum compatibility.
- Quantitative precipitation estimation, nowcasting, radar data processing.
- Tower structure, radome, wind load, electrical resilience.
- Cost, lifecycle, staffing.
This toolkit covers roughly 15% of a full siting study.
Status
Early scaffolding. Per the project plan, the analysis engine (core/) was
developed and validated in a notebook first — the critical path for the
originating Jamaica field report — with core/ and processing/ (the QGIS
plugin wrapper: Evaluate, Discover, Network, Robustness matrix) following
afterward. All four Processing algorithms are implemented; QGIS Plugin
Manager packaging/publication has not happened yet (metadata.txt:
experimental=True).
See Plan_Tecnico_Radar_Siting_Toolkit.md (repository root) for the full
technical plan, phase breakdown, and the reasoning behind each design
decision referenced throughout this documentation.
Testing
tests/ covers every core/ module (pytest, configured via pytest.ini
with pythonpath = .). notebooks/fase0_motor_horizonte_bech.ipynb
validates the engine against wradlib on the real Jamaica radar site and
DEM data (Nivel 2 validation, section 7 of the plan) — wradlib is a
dev/notebook-only dependency (requirements-dev.txt), never a plugin
runtime dependency.
pip install -r requirements.txt -r requirements-dev.txt
pytest
License
GPL v2 or later (LICENSE) — the de facto requirement for QGIS plugins.