REGISTRY
Every index, with its recipe attached.
A public catalog of climate-index datasets, methods, lineage and validation evidence — ready to compare, reproduce and cite.
What every entry carries
Metadata usually buried in papers, notebooks and file attributes — made a first-class part of the result.
SPI · 12-month
Regression covered
Earthflux's reading of the library's wording for library 3.2.0. Not a library claim. The verbatim wording, tolerance and known gap are one step away.
Inspect the example's evidence
- Verbatim library status
- Validated (internal) + qualified external comparison (Pearson III)
- Verbatim library tolerance
- Gamma `atol=1e-8`, Pearson `atol=1e-5` in xarray equivalence tests; NOAA NCEI climdiv characterization ceilings documented in `tests/test_ncei_spi_reference.py`
- Verbatim library known gap
- `tests/fixture/ncei_spi/provenance.json` records that NCEI's climate-divisional SPI product's independence from this codebase's lineage could not be confirmed as airtight from public documentation alone (the survey exists only on the unmerged branch `research/spi-dataset-survey`, commit `79fec847`; the path is absent from `main` and from `v3.1.0`), and that NCEI's actual calibration behavior is a full/expanding period-of-record window rather than its documented fixed 1931-1990 baseline (Baldwin & Chen 2020). WMO SPI User Guide worked-example extraction remains outstanding (issue #778).
Recipe export is unavailable for this illustrative record. It is not a computed result.
Open the stored Lab resultThe Lab opens the stored ACF result, not this example.
Four kinds of evidence
A specification's badge is the weakest Evidence Kind among its index, Distribution and applicable PET Method. A missing component evidence row makes it Experimental.
These definitions are Earthflux's interpretation of the library's wording for climate_indices 3.2.0, not library claims. James Adams approved the nine-member mapping and the Registry Energy mapping on 2026-11-04.
Implementation parity
Same-author reference reproduction, independent-implementation comparison
Library 3.2.0 evidence and limits (d77d3922)Regression covered
Regression coverage, reference reproduction without independence, synthetic characterization
Library 3.2.0 evidence and limits (d77d3922)Experimental
No adopted oracle, specification-level evidence, beta routes, synthetic characterization
Library 3.2.0 evidence and limits (d77d3922)
No stored demo member is Externally validated at 3.2.0. Evidence Kind is separate from origin, reproduction, Reference Checks and Domain Review. The Lab remains a preview with Domain Review pending.
Browse by hazard
Matching entries: 26
Drought
SPI, SPEI, Palmer indices and PET methods — the library’s original core.
Standardized Precipitation IndexSPI
Standardized Precipitation Evapotranspiration IndexSPEI
Self-calibrated Palmer Drought Severity IndexscPDSI
Palmer Drought Severity Index familyPDSI, PHDI, PMDI, Z-Index
Evaporative Demand Drought IndexEDDI
Potential evapotranspiration, ThornthwaitePET
Potential evapotranspiration, HargreavesPET
Reference evapotranspiration, Penman–MonteithETo
Percentage of normal precipitationPNP
Precipitation Concentration IndexPCI
Wildfire
Fire-weather indices on the same gridded inputs and provenance model.
Keetch–Byram Drought IndexKBDI
Canadian Forest Fire Weather Index SystemFFMC, DMC, DC, ISI, BUI, FWI, DSR
Fosberg Fire Weather IndexFFWI
Hot-Dry-Windy IndexHDW
Haines IndexHaines
Flood potential
Wet-extreme and flood-potential indicators for screening, not inundation modeling.
Effective precipitationPE
Effective Drought IndexEDI
Flood IndexI_F
Antecedent Precipitation IndexAPI
Energy
Climate-driven energy indices — degree days, wind and solar resource, and standardized energy droughts.
Heating, cooling and energy degree daysHDD, CDD, EDD
Hub-height wind speedWind profile
Wind capacity factorWind CF
PV capacity factor, temperature-deratedPV CF
Standardized Renewable Energy Production IndexSREPI
Standardized demand indexDemand index
Standardized Residual Load IndexSRLI
Cite this dataset
These identifiers describe the stored ACF preview in the Lab, not the illustrative example above. Inspect the selected recipe and provenance before citing a result.
A DOI is planned. No DOI has been assigned here.
Open methods, hosted workflows
Registry reading and queries stay free. Place assessments and their report-ready exports are paid work; they are not offered by this preview.
Open vs. hosted