I was surprised to learn that clouds create the most uncertainty when it comes to forecasting future climates, due to the interaction between clouds and CO2 particles in the air being highly complex. Before, I assumed that Africa would become increasingly dry, and the Sahara would continue to expand. However, I found it fascinating that Africa's future climate can become either wetter or drier depending on different models and interactions of clouds with CO2. This discussion also made me think about how clouds, which are distributed differently in the same location over time, can influence and potentially create a bias for calculating grid boxes.
During the interview, Professor Palmer explained that the assumptions for grid boxes are based on the assumption that the fluid being measured is completely homogeneous. However, I wonder if clouds are the biggest or only factor that changes the homogeneity of these grid boxes. Is there a way to account for different sheets of grid boxes, corresponding to different altitudes, to further homogenize different portions of the sky? Alternatively, is it too challenging to measure these differences with the current technology we have available?
I was also intrigued by the fact that clouds (which seem relatively simple from an outsider's perspective) can be deeply complex. I appreciated Prof. Palmer's more intuitive explanation of looking out the window during an airport, and upon reflection it makes sense that clouds would have a lot of nonlinear behavior that makes them more complicated to model. The question of how one could vary the mesh sheets in a more adaptive way is an interesting one, but I would bet that Prof. Palmer would challenge it by diving into the complicated question of what standard(s) is best to model the mesh off of, and whether there is overlap between them or if they make radically different setups.