Author: Emma Follis
Data source: NASA GISS Surface Temperature Analysis (GISTEMP v4)
Tools: Python, pandas, numpy, scipy, matplotlib, seaborn
Last updated: June 2026
This project analyzes NASA's GISS Surface Temperature dataset - one of the most cited climate records in the world, maintained by NASA's Goddard Institute for Space Studies. The dataset tracks global surface temperature anomalies relative to a 1951–1980 baseline, covering 146 complete years from 1880 to 2025.
A positive anomaly means that year was warmer than the baseline average. A negative anomaly means cooler. The analysis covers four questions:
- How has global temperature changed since 1880?
- Which decades were warmest and how fast is warming accelerating?
- Are all months warming equally or are some seasons changing faster?
- What does the complete month-by-month picture look like since 1880?
Overall trend: Global surface temperature has risen at +0.0832°C per decade since 1880, with an R² of 0.767 and a p-value of 2.35e-47 - statistically indistinguishable from zero probability of occurring by chance.
Warmest years: The 10 warmest years on record are all from 2015-2025. Not one year before 2015 appears in the top 10. The warmest single year is 2024 at +1.28°C above baseline.
Coldest years: The 10 coldest years are all clustered between 1903-1917, with 1909 at -0.49°C being the coldest on record.
Accelerating warming: Each decade since the 1980s has warmed faster than the one before it. The 2020s (6 years of data) are already the warmest decade on record at +1.065°C above baseline - a +0.257°C jump from the 2010s, the largest single-decade acceleration in the record.
Seasonal patterns: All 12 months have warmed between +0.684°C and +0.815°C since 1880-1980. Winter months (January-March) show slightly greater warming than summer months, consistent with the well-documented pattern of polar amplification - cold seasons and high latitudes warming faster than tropical regions.
NASA_GISS_Temperature_Analysis/
├── NASA_GISS_Temperature_Analysis.ipynb # Full analysis notebook
├── temperature_record.png # Visualization 1
├── decade_warming.png # Visualization 2
├── seasonal_breakdown.png # Visualization 3
├── monthly_heatmap.png # Visualization 4
└── README.md # This file
Option 1 — Google Colab (recommended):
Click the Open in Colab badge above. Download the GISTEMP v4 global temperature
data from NASA GISS
and upload it when prompted.
Option 2 — Local:
pip install pandas numpy scipy matplotlib seaborn
jupyter notebook NASA_GISS_Temperature_Analysis.ipynbDownload the data CSV from the NASA GISS link above.
- Dataset: NASA GISS Surface Temperature Analysis (GISTEMP v4)
- Temperature anomalies relative to 1951-1980 baseline
- Retrieved: June 2026
- 146 complete years (1880-2025) plus partial 2026 data
- Missing values marked as *** in source file - handled via pandas cleaning
- 2026 excluded from annual analysis as the year is incomplete
This project demonstrates end-to-end data analysis in Python including data cleaning with pandas, linear regression with scipy, rolling averages with numpy, period comparison analysis, and four distinct visualization types including a heatmap. The dataset required real cleaning work (handling NASA's *** missing value markers and non-standard number formatting) before any analysis was possible.
Built as part of a scientific data analysis portfolio by Emma Follis, a data analyst with a background in Earth and Planetary Science. Currently completing an M.S. in Space Studies at American Public University while contributing to NASA's Exoplanet Watch citizen science program.