Climate Hazards Center (CHC) skills for SubC multi-model-ensemble forecasts,
Indian Ocean Dipole (IOD) analysis, CPC MJO forecast diagrams, and Africa ITF
position figures. Built on
weather-skills-core
(pinned to main).
| Skill | Role |
|---|---|
subc-mme-fetch |
Fetch and stitch SubC global MME mean + anomaly NetCDFs (7/15/30-day leads) into one forecast Zarr |
iod-mode-index |
Dipole Mode Index (West − East) plus west/east box means from a pre-computed temperature anomaly |
mjo-forecast-fetch |
Fetch the latest CPC CLIVAR MJO Wheeler–Hendon phase-space PNG (GEFS / ECMWF / ECMWF extended-range) |
africa-itf |
Fetch the latest NOAA/CPC Africa ITF / ITCZ position PNG for one region (--location africa|west-africa|east-africa; CPC uses both terms) |
uv sync --group dev
uv run pytest
# SubC MME for one init
uv run skills/subc-mme-fetch/scripts/fetch.py --date 2025-12-01 -o /tmp/subc.zarr
# IOD from a temperature-anomaly field (forecast or observations)
uv run skills/iod-mode-index/scripts/iod.py \
-i /tmp/subc.zarr -v ts_anomaly -o /tmp/iod.zarr
# Latest CPC MJO phase-space diagram (GEFS, bias-corrected by default)
uv run skills/mjo-forecast-fetch/scripts/fetch.py --model gefs -o /tmp/mjo.png
# Latest CPC Africa ITF position map (default --location africa)
uv run skills/africa-itf/scripts/fetch.py -o /tmp/itf_africa.pngIf anomalies are not already in the input, compute them first (climatology +
difference from forecasting-skills), then run iod-mode-index. To view
SST or precip over the Indian Ocean without computing the index, plot those
variables over an Indian Ocean bbox and optionally overlay the west/east
dipole boxes (see iod-mode-index SKILL.md).
./install_agent.sh
# or:
claude plugin marketplace add rhiza-research/chc-skills
claude plugin install rhiza-chc@chc-skillsSame packaging model as
forecasting-skills:
canonical skills under skills/<name>/, Claude plugin via .claude-plugin/ +
agents/, CLI via skills-runner.
See CONTRIBUTING.md.