There's a strong argument for a causal effect (correlation != causation notwithstanding). They showed that temperature variation (colder or hotter than usual)

There's a strong argument for a causal effect (correlation != causation notwithstanding). They showed that temperature variation (colder or hotter than usual) is correlated with a greater volume of tweets overall and also a greater proportion of tweets being hate speech. It seems to me that an increase in the proportion of hate speech could be caused simply by an increase in volume and only have an indirect cause by temperature. I looked but couldn't find anything that might show that's not the case.