Weather Station Analysis

A Python tool for analyzing historical weather data from GHCN stations, creating temperature statistics and interactive visualizations.

Overview

This script downloads weather data from AWS and calculates:

• All-time temperature records for each calendar day
• 1991-2020 normal period averages
• Interactive plots comparing records, averages, and actual data

Data Source

NOAA Global Historical Climatology Network Daily (GHCN-D)

• DOI: 10.7289/V5D21VHZ
• Hosted on AWS S3: s3://noaa-ghcn-pds/csv/by_station/
• Global coverage of daily weather observations

Requirements

pip install pandas matplotlib numpy bokeh

Usage

1. Open HW3_weather_analysis.ipynb in Jupyter
2. Run all cells to analyze:
   • Champaign, IL (USC00118740)
   • Berlin, Germany (GM00010393)

Custom Stations

# Find and analyze any GHCN station
stats = analyze_weather_station('STATION_ID')

Output

• Matplotlib plots: Static temperature analysis with record ranges and actual data
• Bokeh plots: Interactive visualizations with zoom, pan, and hover tooltips
• Temperature statistics: Daily records and 1991-2020 averages for each calendar day

Author

Stephen Allen

ATMS 523 - University of Illinois