Thanks for writing this. The impacts of an AMOC collapse on the food-system implications are definitely worth studying. However, as written the post seems to exaggerate the risk somewhat.
1. The post mixes the impacts of an AMOC collapse in isolation (eg the Jackson paper), with its impacts embedded in a warming world (one of the papers you cite estimates the effects may just cancel). The latter seems more relevant here
2. I also agree with the previous comment that you seem overly drawn to the sudden shutdown literature. There’s a lot of literature on this topic, and it’s probably best to start with the IPCC estimate and update from there.
[More specifically, the higher probability papers you cite generally rely on early warning indicators increasing in variance prior to collapse. This assumes 1. The indicators are highly predictive (which is debated) and 2. That increases in variance are associated with imminent tipping points. The IPCC characterizes the latter as “low confidence” because the same metrics also rise in unforced scenarios.]
Again, not saying this isn’t worth studying, just adding some caution. Happy to talk more if helpful.
This is a great post, I learned a lot, thank you!
This is speculative, but for the specific question of why malaria cases haven’t declined since 2015 I find it hard to believe climate change is a leading factor given the other evidence you present. Here are some reasons:
The only counterargument I can think of is if warming is causing a longer malaria season and people are slow to adjust. This should be detectable in the data, assuming we have seasonal malaria case numbers. On the other hand, this argument still suffers from the uniformity of the pattern, as seasonal temperature and humidity cycles are very region specific and they are unlikely to all lengthen at the same time.
AR6 doesn't comment on autocorrelation (see section 9.2), but the paper actually identifies the variance as a stronger signal: "Here we establish such a measure of the confidence for the variance and autocorrelation and demonstrate that variance is the more reliable of the two."