If you don’t know me, I’m a Ph.D. student studying how clouds affect Earth’s climate. The urgency of climate change makes us forget that the concept of climate itself is a deep and interesting thing. I’d study it whether or not the Earth was warming. This is the first of a series in that spirit: what the climate is, for its own sake.
Today I wanted to show, and explain, a visualization of the difference between weather and climate:
The left is a standard satellite image (available here) showing the cloud patterns for some particular day. The right represents the average cloud patterns over all days: the climate.
The idea for this visualization started when I was looking at a lot of satellite images of clouds for my research on cloud climatology. Everything I was looking at was like the left image—not climate—so the goal felt somewhat abstract.
So, I thought, why don't I just average a bunch of satellite images together?1 To see the process more clearly, take a look at the image below. The top left is again a single image taken at a given time. The top right is the same image, but averaged with one other image from a different day: a two-day average. If you look closely, you can see the same storm systems on the left and right, but on the right they are superimposed on additional storm systems from another day.
The bottom left is 8 images averaged. Individual clouds are beginning to fade, and we can see patterns emerging. The last is what I showed above: 1176 images, taken over a period of 9 years. It’s also cool to see this progression in video form:
Here are average images for two different satellites, showing two different perspectives of North America:
You can learn a lot just by staring at these images. That horizontal bright band near the middle is called the Intertropical Convergence Zone, a region of intense precipitation and clouds. This band forms because the Equator receives more sunlight and therefore more energy, which causes rising air and thunderstorms. Many smaller features are also interesting, for example if we zoom in on Hawaii:
Notice how, just to the west (left) of the islands, it is darker? That is because the wind primarily flows from east to west, and when the air hits the mountains it is forced upward, cools, and forms clouds and precipitation. There is then less moisture available after the air mass passes over the mountains, and so less clouds and a darker average image. You can even see that the islands themselves are whiter on the east side. From that, you might guess the east side is wetter, which is correct.
There’s also a bright spot, dead center in the image, that is farther out to the west of the big island. This is an atmospheric wave forming downstream of Hawai’i, where air is again rising and therefore forming more clouds when compared to the surrounding areas. The effect is similar to the wave that forms behind a speedboat or behind a submerged rock in a stream, but upside down.2
Now that we are familiar with the meaning of the average images, we can be much more creative. All the above images were taken at local noon so that the full disk of the Earth is in daylight. I also combined images from all seasons. But neither is required, because the satellites I am using have a unique orbit that allows them to always stay in the same place.3
Here’s a video where each scene is an average image for a different time of day. The first frame is as above, an average over all images that were taken at local noon. The next frame is for images taken 30 minutes after local noon, the next 1 hour after noon, etc. The result is “an average day on Earth”:
The bright spot that moves across the image is the reflection of the sun in the ocean. Notice how it gets to quite a high latitude during the nighttime hours? That’s because I made this video using only June images—so the pole at the very top never sees night. You can see how far north the sun rises and sets in summer.
By the way, the above image also demonstrates the moon’s crescent shape in a visually intuitive way: it’s just a ball lit from the side. You often can’t see the dark side of the moon,4 so (at least for me) we experience a kind of illusion where the crescent appears flat, like a paper cutout.
We can also fix the time of day but vary the seasons we look at. For the below video, each frame is an average image for all noon images of a given month. One frame is all noon images for January, another for February, etc. An “average year on Earth”:
In this one, we can see the seasons pass. Central America and central/southern South America have particularly notable wet/dry seasons. We can also see the Intertropical Convergence Zone (the bright horizontal band) move north and south, roughly tracking the latitude that receives the most solar energy. Snow cover forms in the winter in Canada and the southern Andes at the south tip of South America.
Here’s a seasonal video using the dawn images rather than noon, which highlights how the sun’s position changes throughout the year:
That’s all I did, but there are probably many more visualizations that are possible, particularly for smaller local regions. I put the code and images on GitHub, including high-res versions, if you would like to download them or create your own. Link here if you use them and let me know if you make something cool!
So that’s climate, visualized. Soon I’ll do another post looking deeper into the idea of climate, so stay tuned for that.
1Literally by taking the numerical RGB values and averaging them across multiple images.
2It’s also not as persistent as a wave behind a speedboat. We have to remember we are looking at an average over years here, and the atmosphere is not going to look like this every day.
3Their orbit is perfectly in sync with the rotation of the Earth, so they appear to stay at a fixed location while also staying in the sky.
4But it is definitely possible. Take a look under a dark sky when the moon is a relatively thin crescent. This will be near dawn or dusk. At this time, the moon “sees” a relatively bright Earth (just not bright where you are standing), and Earth is bright enough to light up the dark side of the moon enough to see it.