CO2 level of air and global temperature, the future
Continuing the previous blog post, I was wondering what about the future? Can I predict the future temperature levels based on generated carbon dioxide levels?
I’m trying to do this in this post.
Assuming despite our actions there is no change in the our carbon dioxide emission we can generate the future levels with a simple addition:
Well, it’s not a very uplifting graph.
And this is the temperature anomaly graph from 1850 to 2018 from my previous post. This is a gridded dataset of global historical surface temperature anomalies relative to a 1961–1990 reference period.
Now I can combine the past and the generated carbon dioxide level data with the temperature data from the past, fit some models and plot the results.
Baseline first, a linear regression model:
As I expected, I got a continuously increasing, underfitting line. Let’s see if I can a little bit more “seasonal” graph.
For this I used the Facebook Prophet model. First I checked if the model can do any better with the CO2 level data:
It’s basically the same.
Now let’s see the temperature anomaly data:
This graph shows a little bit more the seasonality and looks more like the past temperature anomaly graph. For this prediction the mean absolute error is 0.5399.
And now a little information about what inspired this post:
There is a youtuber named Mr. Beast. He reached 20 million followers on Youtube a couple months ago. He usually celebrates these milestones with some challenge and a big giveaway. The 20 million was a little different: lots of his followers commented instead of a giveaway he should plant 20 million trees. Well, he listened: he joined forces with teamtrees.org, and they trying to plant 20 million trees before 2020! This made me thinking: what is the effect of 20 million trees to the carbon dioxide levels of the atmosphere? Is there an effect? What change 20 million trees will make to the temperature?
A little math: to directly compare CO2 emissions to atmospheric CO2 levels, both sets of data can be converted to gigatonnes of CO2. The CO2 emissions data is typically expressed in gigatonnes carbon (GtC). One gigatonne is equal to one billion tonnes. The atomic mass of carbon is 12, while the atomic mass of CO2 is 44. Therefore, to convert from gigatonnes carbon to gigatonnes of carbon dioxide, simply multiply 44 over 12. In other words, 1 gigatonne of carbon equals 3.67 gigatonnes of carbon dioxide. Atmospheric CO2 levels are expressed in parts per million by volume (ppm). To convert from ppm to gigatonne of carbon, the conversion tables of the Carbon Dioxide Information Analysis Center advise that 1 part per million of atmospheric CO2 is equivalent to 2.13 Gigatonnes Carbon. Using the 44 over 12 rule, this means 1ppm = 7.81 Gigatonnes of Carbon Dioxide.
Trees have this amazing ability of being almost entirely made up of carbon atoms. They need very few nutrients relative to their size so they don’t deplete the topsoil, get most of their water from deep underground, and their mass from the air they breathe. This means that the lifetime carbon value of a tree, the total amount of carbon it will absorb, is about the same as its mass. So what’s the average mass of a tree? About 2 tonnes.
So if we want to remove 1ppm (2.13 billion tonnes of carbon, 7.81 billion tonnes of CO2) of carbon dioxide from the air, we need to plant 1.065 billion trees.
In conclusion: 20 million trees is a good start, but we need more.
How many new Trees would it take to Offset our Carbon Emissions?
Last year in 2018, the world emitted an estimated 37 billion tonnes of Carbon Dioxide. If you compressed all that back…