Oman Climate Atlas

View the live atlas →

A reproducible climate atlas for Oman, built from public weather and sea-surface temperature data. The project starts with Muscat/Seeb and expands across a small station catalog covering coastal, interior, mountain, and khareef-influenced locations.

The public site is a static atlas generated into site/, with no application server required:

https://nasseralbusaidi.github.io/OmanClimate/

What This Shows

• Long-term heat, 30°C-night, heatwave, summer-length, and wet-bulb trends.
• A station comparison across Muscat, Salalah, Sohar, Sur, Nizwa, and Saiq.
• Personal climate comparisons from a selected birth year to the latest full year.
• Story-signal summaries for December cool snaps, Salalah khareef stress, and Saiq as a mountain-refuge comparator.
• Sea of Oman SST context from NOAA OISST, including associations with coastal humid-heat metrics.
• Annual climate outlooks through 2070, with each forecast carrying rolling one-year hindcast accuracy and skill against a persistence baseline.

The project is designed to be legible rather than overconfident. Published trend fits use the trusted year >= 1980 window and drop incomplete years with n_days < 360. Pre-1980 ERA5/Open-Meteo points remain useful context, but they are not used for the public trend lines because diagnostic tests found unstable early-reanalysis behavior in this region.

Data Sources

Source	Use	Output
Open-Meteo Archive API	ERA5-backed hourly temperature, humidity, and dewpoint for Oman stations	data/raw/open-meteo/, data/processed/*.parquet
NOAA NCEI GHCN-Daily	Seeb International station observation cross-check	data/raw/ghcn/, data/processed/muscat_ghcn_*.parquet
NASA POWER	Independent daily gridded weather cross-check for annual temperature signals	data/raw/nasa/, data/processed/*_nasa_*.parquet
NOAA OISST v2.1 via NOAA PSL THREDDS	Sea of Oman monthly sea-surface temperature context	data/raw/noaa-oisst/, data/processed/sea_of_oman_sst_*.parquet

Raw and processed data files are ignored because they are regenerated from the public sources. The committed site/ directory contains the compact static payload used for GitHub Pages.

Repository Layout

pipeline/              Fetching, processing, analysis, diagnostics, and charting
tests/                 Pytest coverage for data processing, analysis, CLI, and site outputs
docs/methodology.md    Detailed methodology, caveats, thresholds, and changelog
site/                  Static atlas artifact published by GitHub Pages
notebooks/             Exploratory source-comparison scripts
data/                  Local regenerated caches and parquet outputs
charts/                Local regenerated chart working directory

Quick Start

Install the project with uv:

make install

Run the complete local data pipeline (fetch, process, and build the static site):

make pipeline

Open the static atlas locally:

open site/index.html

Run the test suite:

make test

Step by step

make pipeline runs these stages in order; each can also be run individually:

make fetch
make fetch-ghcn
make fetch-stations
make fetch-nasa
make fetch-sst
make process
make process-ghcn
make process-stations
make process-nasa
make process-sst
make site

Common Commands

make charts                 # Muscat chart set
make phase3-charts          # Multi-station charts
make station-map-data       # Static station-map payload
make personal-climate-data  # Static personal-comparison payload
make story-metrics-data     # Static story-signal payload
make sst-data               # Static Sea of Oman SST payload
make predict-data           # Static validated climate-outlook payload
make diagnostics            # Window, step-change, and rural-comparator diagnostics
make clean                  # Remove regenerated local data/chart/site artifacts

Methodology

The detailed methodology lives in docs/methodology.md. It documents:

• timezone conversion before daily aggregation,
• wet-bulb calculation and thresholds,
• trend statistics and uncertainty bands,
• the trusted 1980-present trend window,
• station roles and urban-heat-island caveats,
• GHCN, NASA POWER, and OISST source handling,
• statistical prediction validation and current limitations.

Deployment

The repository includes .github/workflows/pages.yml. On pushes to main, the workflow uploads the committed site/ directory to GitHub Pages. It does not run the Python pipeline in CI; regenerate and commit site/ whenever the published atlas data changes.

License

OmanClimate is released under the MIT License. See LICENSE for details.