How I Built a Weekly Comfortable Region Ranking Using Weather Data

#weather #climate #opensource #database

How I Built a Weekly Comfortable Region Ranking Using Weather Data
I have been working on a personal project called weather2, a weather analysis platform that automatically collects weather data and generates multilingual reports.

Recently, I published a historical dataset called:

Weekly Comfortable Region Ranking Dataset

The idea behind the project is simple:

Which cities around the world have the most comfortable outdoor conditions this week?

Instead of looking only at temperature, I combine several weather indicators to calculate a custom Comfort Score.

Metrics Used
The ranking is based on:

Temperature
Humidity
Discomfort Index (DI)
Heat Index
Wet Bulb Temperature
WBGT (Wet Bulb Globe Temperature)
These indicators help estimate how comfortable or stressful outdoor conditions may feel.

Data Collection
Weather data is collected automatically and processed through SQL-based workflows.

The project currently generates:

Multilingual weather reports
Historical weather reports
Weekly Heat Stress Index
Weekly Comfortable Region Ranking
Historical CSV datasets
Dataset
I recently published the complete historical dataset as a CSV file.

Dataset page:

https://yahikoyama.github.io/weather2/comfortable_region_dataset.html

CSV download:

https://yahikoyama.github.io/weather2/CSV/all_weekly_comfortable_region.csv

Example Fields
The dataset contains:

Ranking
FromDate
ToDate
City
Country
Average Temperature
Average Humidity
Heat Index
Wet Bulb Temperature
WBGT
Discomfort Index
Comfort Score
Why I Built It
Many weather sites focus on raw temperature values.

I wanted a way to compare locations using multiple comfort-related indicators and produce an easy-to-understand ranking.

The goal is not to determine the "best" city, but to create a reproducible dataset that allows comparison between locations and climate conditions.

Project Links
Website:

https://yahikoyama.github.io/weather2/

GitHub:

https://github.com/yahikoyama/weather2

Dataset:

https://yahikoyama.github.io/weather2/comfortable_region_dataset.html

Kaggle Dataset:

https://www.kaggle.com/datasets/yahikoyama/weekly-comfortable-region-ranking-dataset

I'd be interested in feedback on:

Comfort score methodology
Additional weather indicators
Dataset structure
Interesting cities to add