
Wie verfolgen wir den Klimawandel?
Von allen Diagrammen auf unserer Website ourworldindata.org reicht dieses so weit zurück in die Zeit wie kaum ein anderes: die CO₂-Konzentration in der Erdatmosphäre in den letzten 800.000 Jahren.
Die meiste Zeit über stieg und fiel der CO₂-Gehalt zwischen etwa 180 und 300 Teilen pro Million (ppm), was auf natürliche Kreisläufe in der Erdumlaufbahn zurückzuführen ist. Wir wissen dies aus Kernen des antarktischen Eises, in denen sich bei der Entstehung winzige Luftbläschen befanden.
Bis zum Jahr 2025 erreichte die Konzentration – die jetzt vom Global Monitoring Laboratory der NOAA gemessen wird – 426 ppm, mehr als jemals zuvor in diesem Rekord. Die Konzentrationen sind im letzten Jahrzehnt jedes Jahr um 2 bis 3 ppm gestiegen.
Dieser dramatische CO₂-Anstieg ist hauptsächlich auf menschliche Aktivitäten zurückzuführen; Diese und andere Treibhausgasemissionen sind die Hauptursachen für den aktuellen Klimawandel.
Die CO₂-Konzentration ist nur einer der Indikatoren, anhand derer wir den Klimawandel verfolgen. Andere umfassen globale Temperaturen, den Wärmegehalt der Ozeane und die Schneedecke. Jedes davon wird von einer speziellen wissenschaftlichen Einrichtung zusammengestellt, vom britischen Met Office Hadley Centre bis zur NOAA in den USA.
Datenquelle: NOAA Global Monitoring Laboratory (2026); US-EPA basierend auf verschiedenen Quellen (2022)
Verwendete Werkzeuge: OWID-Grapher und Figma
Von ourworldindata
24 Kommentare
**Data source:** [NOAA Global Monitoring Laboratory](https://gml.noaa.gov/ccgg/trends/gl_data.html) (2026); [US EPA](https://19january2025snapshot.epa.gov/climate-indicators/climate-change-indicators-atmospheric-concentrations-greenhouse-gases/index.html) based on various sources (2022)
**Tools used:** OWID-Grapher and Figma
Why was the cutoff point 800,000 years ago?
Can we go further back please! Like I know 800,000 is ages ago but would be good to see millions of years as well
Can you please change the year to a reverse log scale
Edit: I believe reverse log is the wrong term to use in this case but I hope you understand what I mean.
that’s one hell of a X-axis scale
Okay but we didnt have intelligent species back then as well? You are comparing alot of different erasThis is meaningless.
Seems very anti ai propaganda.
I’m sure private jets, factories, china and India are doing way more damage than what was happening 800,000 years ago, yet it seems everything was normal until 2025
Edit… Should say early 21st century.
Industrialization hits hard
This is the chart you show people that say “earth has natural heating and cooling periods” yeah dipshit it does, that’s the noise at the bottom of the graph
With that scale, it looks like all the CO2 happened after 2025
Not contradicting the truth, but the graphs should start at 0 ppm. This is misleading. It looks like 426 is double the 300 ppm
Elon Musk says 1000ppm is a good target to aim for
Why are there uniform spikes? What are they?
Yes, we have a problem with climate change. However, I hate this kind of chart because we don’t have 800,000 data points for each year. These spikes in the past could have exceeded 300 ppm, but were flattened by multiple years of data compressed into one datapoint.
Why can’t y start at zero… It’s bad but need to show it correctly
How confident are we in our ability to measure transient, relatively short duration spikes, in this historical peaks? Eg is it possible the true spikes were 400+ ppm but happened so quickly (on an 800,000 year scale) that we couldn’t measure it?
Or is measuring CO2 ppm like counting growth rings on trees which we can do with extreme precision?
So the next ice age is cancelled then?
Would love to see what the spike looked like during events like the Siberian Traps volcanic eruption
It’d be interesting to test if the dramatic rise in co2 is a contributing factor to increasing anxiety disorders
The x-axis is a bit misleading at first glance. It almost looks like the spike starts after 2025.
Would love to see this with 0, 50, and 100 shown.
We definitely need more data centers
Should be called CO8 now then!