Monthly Temperature Anomalies: an Interactive Data Story from Our World in Data
This dataset gives the difference between each month’s average surface air temperature and the 1991–2020 average for the same month. It comes from the Copernicus ERA5 reanalysis and covers every country and region month by month since 1940.
On this page the dataset is downloaded live from Our World in Data, straight into your browser, and turned into an interactive dashboard and a data story told as a tap-through stories. Nothing is uploaded to a server and nothing needs installing.
Interactive dashboard and data story
At a glance
| Indicator | Monthly Temperature Anomalies |
| Unit | °C |
| Coverage | monthly values from 1940 to the latest release |
| Highlighted entities | World, Europe, Asia, Africa, North America |
| OWID chart id | monthly-temperature-anomalies |
| Column | temperature_anomaly |
| Story format | Tap-through stories |
| Source | Our World in Data, CC BY 4.0 |
Why this dataset is told as a tap-through stories
The warming stripes work best full screen, one screen at a time. Tap-through stories give the data its immediate impact.
Phone-sized cards with progress bars, the way social stories work. Tap right for the next card, left to go back. Datastripes can tell any dataset in eight formats: scrolling article, slide deck, conversation, newspaper front page, timeline, tap-through stories, quiz and bento grid. In automatic mode it picks one from the shape of the data. Here the format is fixed to the one that fits this indicator best. You can see every format side by side in the data story gallery.
Background
Anomalies remove the seasons, so a positive value means a month was warmer than usual for that time of year. For most of the mid-20th century anomalies hovered around zero against an earlier baseline; since the 1980s they have risen steadily.
The years from 2023 to 2025 were the warmest on record, with many individual months more than 1.5 °C above pre-industrial levels. Land regions, and Europe in particular, are warming faster than the global average.
How the indicator is measured
ERA5 is a reanalysis: a weather model is run over the past and continuously corrected with billions of observations from weather stations, balloons, aircraft, ships and satellites. Anomalies are computed against the 1991–2020 monthly averages.
What to look for in the data
- Recent anomalies are the highest in the whole series.
- Europe shows larger swings and a faster warming trend than the world average.
- El Niño years stand out as global spikes, for example 1998, 2016 and 2023–24.
- Negative anomalies have become rare over the last two decades.
Limits and caveats
- Because the baseline is 1991–2020, early decades look colder than against a pre-industrial baseline.
- Single-country monthly series are noisy; trends are clearer over years and regions.
- Reanalysis is less constrained in early decades and in regions with few observations.
Who uses monthly temperature anomalies data
- Climate scientists and students
- Weather and climate journalists
- Insurance and risk analysts
- Agricultural planners
How to analyze this dataset yourself
- Open the data. Download the CSV from the Our World in Data chart and drop it into Datastripes. It loads in your browser, with no upload.
- Filter the entities. The file has one row per country and month. Filter the
entitycolumn to the countries or regions you care about, or usecode(for exampleera5) to pick out aggregates. - Pick a time window. Sort or filter on the time column to isolate a period, such as the years before and after a policy change.
- Join other indicators. Add a second OWID file, for example GDP or population, and join it on entity and year to look for relations.
- Forecast and simulate. Run a no-code forecast on any series, or build what-if scenarios on top of it.
- Publish a story. Turn your filtered table into a data story in any of the eight formats and embed it on your site with the Datastripes dashboard widget.
Related data stories
- CO2 Emissions: timeline
- Renewable Energy Share: slide deck
- Energy Consumption: conversation
- CO2 Emissions per Capita: scrolling article
- All data stories in the gallery
Frequently asked questions
What is a temperature anomaly?
The difference between a measured temperature and a long-term average for the same place and time of year.
Why 1991–2020 as a baseline?
It is the current standard climate normal of the World Meteorological Organization.
Is this the same as 1.5 °C above pre-industrial?
No. The Paris targets use an 1850–1900 baseline, which is about 0.9 °C cooler than 1991–2020.
Data: Our World in Data, published under CC BY 4.0, downloaded in your browser when you open this page. Datastripes is not affiliated with Our World in Data. Visualisations are generated automatically and may lag behind the latest OWID release by a few days.