About & credits
The project, its sources and how to use or contribute to the material
About the project
PyCPT Tutorial connects readable climate-forecast stories with executed notebooks and traceable data to teach the use of PyCPT. CPT is the underlying engine for PyCPT model fitting; this is not a standalone CPT installation or operation guide. Its first case examines seasonal rainfall forecasts over Indonesia across twelve windows centred on 2023. Seasonal and subseasonal topics are part of the project’s broader scope; only completed, verified material is presented as an available case.
The current case is complete as a reproducible tutorial. Physical interpretation of fitted patterns, longer independent validation and probability reliability remain separate research questions. See methods and limitations for the evidence supporting the published comparisons.
The source is maintained in the pycpt-tutorial repository. This is a tutorial project, rather than a replacement for upstream software documentation.
Upstream software and teaching references
- PyCPT source and documentation
- IRI PyCPT example notebooks
- IRI teaching reference: CPT, calibration and cross-validation
- CPT help
- Quarto, used to publish the story, notebook and guides
The executed case and its recorded configuration determine the behavior described here. The reader guide separates browser reading, offline analysis, optional PyCPT fitting and conditional data-access requirements.
Data and boundary provenance
The case uses CFSv2 model rainfall and CHIRPS3 observations. Country and province boundary files supplied for the project provide map context and regional assignments. The analysis and simplified display boundaries have separate provenance records.
| Material | Source record |
|---|---|
| Packaged case inputs, calendars and hashes | Data provenance |
| Country, province and neighbouring-country map overlays | Display-boundary provenance |
| Native province assignments, support and residual | Province-analysis provenance |
| Browser-readable chart and map inputs | Interactive-data provenance |
Citation and reuse
When referring to this work, identify PyCPT Tutorial, the case title Twelve seasons over Indonesia, the repository revision you used and your access date. Credit the upstream software and underlying data sources as appropriate. The source repository records revisions.
The repository includes a Mozilla Public License 2.0 license file. Upstream software and third-party data or boundaries have their own source terms; the provenance links above identify the supplied material. No additional license is introduced by this website layout.
Questions, corrections and contributions
Report an issue in the repository. Include the page or notebook cell, the selected season/metric/period if relevant, and what you expected to see. For execution problems, include the software versions and a concise error message; keep credentials and private paths out of public reports.
For an explanation of the story/notebook approach, begin with the reader guide. For runnable inputs and environment instructions, use setup and downloads.