As a "cloud person", I just want to add a few things to the description of how clouds affect the climate (and why high clouds have a wamring effect).
All clouds are white, so they all reflect sunlight back into space (during the day), cooling the Earth.
All clouds are (almost) black in the infra-red, meaning the amount of energy they emit in the infra-red is determined by their temperature. Colder clouds emit less energy.
Almost all clouds are colder than the surface beneath them, which means they emit less infra-red energy to space than a clear day would. This reduces the amount of energy the Earth emits to space, so warming the climate.
High clouds are colder than low clouds, so have a stronger warming effect.
I have a question for a cloud person, maybe you can answer it.
When it rains, where does the latent heat go? The latent heat of evaporation (or condensation) is absolutely huge. Condensation means heat is released. I did a back of the envelope calculation. 2 mm daily rainfall x 500 million km2 = 10^15 kg; each kg of water holds 2.26 MJ of latent heat, and there are 86400 seconds in a day, so that's 26.15 W, so overall 26150 TW. The Earth receives about 173000 TW from the Sun, so this is about 15% of the energy received from the Sun. Obviously, not all the 15% goes out to space, but about how much does go to space?
It heats up the atmosphere and eventually gets emitted back into space!
For the Earth's temperature to remain approximately constant, the energy leaving the system (as infra-red) has to balance the energy entering the system (as sunlight).
The atmosphere is almost transparent to visible light, so sunlight doesn't really heat the atmosphere at all, it mostly heats the surface.
In contrast, the atmosphere is mostly opaque to infra-red (apart from the 'window region' at about 10um), which means energy is mostly emitted from higher levels in the atmosphere.
This means that you have to have a way of getting energy from the surface (were it effectively 'arrives') to higher levels in the atmosphere (where it can leave the Earth system again. Latent heat is an important way for this to happen - you can see it in this figure, showing how energy flows in the Earth system
I didn’t install the plug-in, but made a mental substitution and reread the thread. Now I am chuckling like a grade schooler over some madlibs and dribbled coffee down my shirt.
Cloud computing energy use appears to be on an exponential trend driven by general trends (all things automated), with new forms of automation compounding competitive pressures (deep learning models quickly getting larger, more powerful, more useful, and more versatile in a way erasing many lines holding back past competition.)
At some point, it seems inevitable that computing usage will be a first level climate driver, regardless of how green the energy is.
Harnessing orbital solar, fission and fusion power, may solve the CO2 energy problem, without requiring us to steal the biosphere's energy needs, but will eventually create a massive waste heat energy problem.
Unless we find someway to efficiently transfer mass amounts of heat energy off of Earth.
Or we eventually limit computing on Earth, and export that to the Moon and beyond.
Not all water vapor is clouds. Where relative humidity is too low it comes with all the warming but none of the cooling effect. A tiny change in humidity (cloud/no cloud) can change things significantly.
This IR satellite view of clouds and water vapor centered over N/S America is pretty nice. You can also see the daily pulse of cloud formation over the Amazon rainforest:
That's interesting. How relevant is energy transport in comparison to the optical effects?
Eg: my understanding is that hurricanes are net cooling because they transport heat from the ocean surface to the upper atmosphere. Presumably the same can be said for cumulonimbus/thunderheads? Or perhaps it is more relevant when they form in the day and when they dissipate at night?
This is related to my reply above, but clouds in general move heat upwards in the atmosphere through latent heating.
When you evaporate water from the surface, you cool it (like sweating keeps you cool). This water vapour is then lifted by convection until it cools enough to condense and form a cloud. As the water vapour condenses, the opposite happens and it heats the atmosphere locally (this further invigorates the convection)
Once you have condensed enough water (and the water droplets/crystals are large enough), you form precipitation. This falls back to the surface (some evaporates along the way), where the process starts again.
This transporting of energy through the water cycle is an important component of how energy moves in the Earth system - you can see it on this figure as 'latent heating', moving energy away from the surface at something like 80Wm^-2
I will give you the benefit of doubt given it's Hacker News you likely are an expert, but this feels like one of those "sounds too intuitive to be that simple" type complex factors. Any literature on the topic from which I can improve my understand?
What part of it do you doubt? It’s obvious that clouds are white, and it’s obvious that they are cooler than the earth and even colder higher up (clouds form because the adiabatic expansion of rising air cools it down and causes water to condense, and the higher up you go the colder the atmosphere due to greater expansion - anybody who has hiked the mountains has experienced this).
The only assertion here that one has to take on faith is that clouds are approximate black bodies at infrared wavelengths (which isn’t surprising - most things tend to be), and the relative magnitude of the cooling vs warming effects. Oh and there is an unstated dependency that the Earth is also an approximate black body at infrared wavelengths.
I think the concept of the albedo of the poles is another simple idea that explains why melting poles would be bad and accelerate global warming further, because by being white they reflect back sunlight for a huge area. Kinda intuitive if you have ever touched the black (opposite color) coolers on a desktop computer and felt your hand fry
Granting that the two effects counteract each other, it is surprising that the crossover point would happen within the range of variation of actual clouds on Earth. My uninformed guess would have been that nearly all clouds' effects are dominated by only one of the two effects—probably, by the cooling effect.
All clouds are white, so they all reflect sunlight back into space (during the day), cooling the Earth.
All clouds are (almost) black in the infra-red, meaning the amount of energy they emit in the infra-red is determined by their temperature. Colder clouds emit less energy.
Almost all clouds are colder than the surface beneath them, which means they emit less infra-red energy to space than a clear day would. This reduces the amount of energy the Earth emits to space, so warming the climate.
High clouds are colder than low clouds, so have a stronger warming effect.
In summary:
Low clouds - Reflect sunlight (cooling), don't trap much infra-red (little warming)- Net: Cooling effect
High clouds - Reflect sunlight (cooling), trap lots of infra-red (stronger warming) - Net: Warming effect