> Finally some science, ok cool. So we're basing this on the baseline of no contrails what-so-ever. We measure the radiative photons? of the light refracting? through it and see it's long-wave radiation channeling properties scatter it rather than allow it through to relieve the earth's heat?
There are many different avenues you could pursue to estimate this, with different degrees of complexity and they all boil down to a radiative transfer calculation. You could start with an idealized, single-column model using a very simple approximation like a 2-stream approach, crudely parameterizing how a few bulk properties of a cloud in the column would influence the estimate of radiative balance at the top of the atmosphere. Such a model would let you explore a huge variety of factors.
To get more into the world of "real climate", you'd eventually scale up to an atmosphere model with a more complex 3D radiative transfer scheme, probably with more complex cloud microphysical representation and interactions with radiation. This still lets you do idealized experiments.
The next step up, you'd do what the DeepMind and Contrails teams have done, which is to use a standard Earth System Model ("climate model"). That just builds the complexity, and allows you to study realistic scenarios with counterfactuals (e.g. with or without additional contrail formation due to aviation). It's still going to be parameterized in some sense.
Ultimately, we'd look at real-world data top-of-atmosphere radiative estimates, which we conveniently measure from a variety of satellites (CERES would be the most well-known one). The problem is that we can't study the counter-factual very easily in the real world. We get a few isolated cases (like the aviation pause over CONUS after 9/11, or the global decrease in aviation during COVID), but there's so much inherent noise (from weather) that it's hard to infer any causality; you'd likely end up using that climate model from the last part to simulate the counterfactual in a weather scenario nudged to reproduce the real-world observations you have.
This would make a great Masters thesis, or maybe a slice of a PhD, but it's not much more complex than that.
> As for contrails, if they're likely to form in certain areas and not others, what's that coefficient? is it the moisture content? is it the jet fuel mixture? Since you say the Contrails program helps predict when these might occur, is that data being used to alter cruise paths?
It's the ambient thermodynamic and kinematic parameters. Namely, background water vapor and temperature, combined with local kinetic energy in the form of turbulence that you need to drive very local supersaturation. Given those parameters, different distributions of gases and particulates emitted from a jet engine may be more or less effective at nucleating ice - if the conditions are even right to do so.
It's worth pointing out that it's _very_ hard to capture these background conditions in weather models, hence why a major part of the DeepMind/Contrails project have been to use machine learning to try to constrain background circulation patterns most consistent with contrail develop as evidenced by high-resolution remote sensing and imagery. Some of the earliest work they did was to use these types of tools to generate data that allow us to better parameterize contrail formation in numerical weather/climate models, which is why they're able to turn this into a predictive and actionable problem.
Otherwise, there's an 80-year literature on this topic. I encourage you to read it. Kärcher's review paper in Nature in 2018 is a good starting point (https://www.nature.com/articles/s41467-018-04068-0).
> Any route (with the exception of the SIDS and STARS) is charted across airways to the most direct route possible. Right?
I'm not an expert in aviation route planning, but from personal experience I know that many other factors come into play with routing (namely avoiding weather, limiting time spent extremely remote / far away from airport or other infrastructure, traffic / network congestion, etc).