But I feel we also run into the reverse problem. While it's true that neither Hurricane Irma nor Harvey nor Katrina, nor the Californian or Australian wildfires, nor the strangely warm arctic temperatures, can be singularly attributed to climate change, the pattern of all these events is clearly climate change at work.
And so we end up stuck in this stupid limbo, where because no one can be certain that climate change caused this disaster, nothing is ever done in response to any disaster. We should be viewing these events as a serious call to action.
> ...the pattern of all these events is clearly climate change at work.
This is still debated. The notion that weather has become more extreme still relies on very little data and statistical models trying to fit different types of events into the same model. ...and none of the models are able to digest all types of events across the entire world, over the entire 100 year reference time window.
But more importantly, the debate around it is a distraction.
The proven fact is that global warming has occurred and continues, AND that it will cause sea level rise within our lifetime. That fact has sufficient serious consequences for nations/states/cities to begin planning local responses, and maybe coordinate global emissions reductions (though this seems increasingly too little, too late).
Good call out. Not just a disservice and unscientific, it’s just plain wrong. An observed variance in weather over one day or even a few days isn’t even climate. It’s just weather activity.
Sorry, but increased variance is very much in line with systemic climate change. In fact, it's one of the main predictions as climate change means disruption of the normal distribution, not just pushing the median but increasing the standard deviation.
> increased variance is very much in line with systemic climate change
Not really, it's also in line with the limited amount of high-quality data we have.
E.g. viewed in isolation, recent Gulf of Mexico hurricanes were pretty bad (2019 in particular, IIRC). But if you look into history, 1932,1933 was also pretty bad.
Arguably, if we had even more recorded history, such outliers wouldn't seem that exceptional - there still is a possiblity that climate change is affecting hurricanes (making them worse) but you can't just look at the variance (outliers) - trend is also (more?) important.
Let's use math and statistics, not anecdotes. If an increase in variance means that you get a +5 outlier 10x more often, then looking at a 10x longer timescale will mean that you see the +5 outlier roughly as often, making it seem "less exceptional". The problem is that the exceptional thing is now happening 10x more often.
For example, Texas getting three "500 year" floods in 3 years. These were floods that were estimated to be a 0.2% probability in any given year. The fact that they are happening at much higher frequency is because the entire probability distribution has shifted (and expanded). Now if you improved your ability to peer into the past and found another "500 year flood" a few hundred years back, you'd be tempted to think everything is fine, when everything is decidedly not fine.
You cannot take prove a global phenomenon with local data, unless you compare global occurrences of local events.
Taking only Texas as an "example" means you're cherry picking data. I'm sure we could find other parts of the world that had multiple rare events in every decade of the last century as well.
It was just an example. If you look at all the data together, the trend is clear. Both median temperatures and variance are increasing. Same for other extreme weather events like hurricanes, wildfires, and droughts.
Is there a data source that says that temperature variances have increased? I have never seen meaningful evidence of this.
> Same for other extreme weather events like hurricanes, wildfires, and droughts.
Extreme weather events are extremely hard to define, and have such a low occurrence rate that you cannot extrapolate an increase in the timeframe we're discussing on a global level.
> Arguably, if we had even more recorded history, such outliers wouldn't seem that exceptional
This holds true if you assume a system is static, which is one of the hardest communication points the scientific community has to convey to the general public. A good part of climate science is not just mapping the statistics of a system, but analyzing those in terms of physics and chemistry to figure out what causes dynamism in the climate system.
The driving mechanic behind climate change theories is that we are changing the chemistry of the atmosphere, which is making the climate unstable. If you just look at the statistics through the lens of data science, it's easy to write off effects as "not enough data", but this is a simplistic view, stemming from not spending huge amounts of time learning and studying climate science.
> The driving mechanic behind climate change theories is that we are changing the chemistry of the atmosphere, which is making the climate unstable.
This is a strange characterization of the increase in CO2 and resultant heat retention. It's not itself "unstable".
> If you just look at the statistics through the lens of data science, it's easy to write off effects as "not enough data", but this is a simplistic view,
Selecting only a specific set of events in specific areas is, by definition, cherry picking data.
> Selecting only a specific set of events in specific areas is, by definition, cherry picking data.
You used the same phrase elsewhere in the thread. I am not sure if we agree on what "cherry picking" really means. There is a difference between isolating relevant variables and cherry picking data. For example, if you wanted to track the long term trend of lion weight in Africa, you wouldn't also include in your dataset the weights of monkeys in South America. That'd just add irrelevant noise because, even though they are measurements of animal weight, they are not what we are studying. It absolutely would not be "cherry picking" to exclude monkeys from your sample. Real cherry picking would be things like excluding particular years of measurements or finding other reasons to exclude data that goes against a trend you are looking for.
> This is a strange characterization of the increase in CO2 and resultant heat retention. It's not itself "unstable".
I'm not sure how you got this from what I wrote. I specifically said "changing the chemistry of the atmosphere", which is not a strange characterization of an "increase of CO2" in the atmosphere, it is word for word a correct description of that action. I have no idea what you mean by "It's not itself "unstable", especially in the context. If you mean that changing the chemical makeup of the atmosphere doesn't make climate patterns unstable, then you're just wrong.
> Selecting only a specific set of events in specific areas is, by definition, cherry picking data.
This also literally has nothing to do with what I said. Are you some kind of weird GPT bot or something? I'd assume you accidentally wrote back to the wrong comment, but you quoted me... If you'd go ahead and reread my post, and not skip the big words, you'll see that I was trying to convey that because we have such limited data, and it's data from a system that's ridiculously large and dynamic compared to the data, any purely statistical analysis is going to be absolute garbage. Hence, you also need to use logical reasoning about cause and effect, what types of chemical shifts in the atmosphere will cause what types of outcomes.
But really, you need to stay a little more on task if you'd like to discuss and debate. Just confusing at this point, and I'm still not sure if you're trolling or a bot or just flailing.
If you observe a few days of severe weather, you've observed severe weather. Taking a mere observation over one week of weather activity cannot be looked at in isolation and simply called "climate change, tada!"
FTA:
>When scientists talk about climate, they're often looking at averages of precipitation, temperature, humidity, sunshine, wind, and other measures of weather that occur over a long period in a particular place. In some instances, they might look at these averages over 30 years. And, we refer to these three-decade averages of weather observations as Climate Normals.
There is, which is why it's a disservice -- and unscientific -- to say things like: it's colder than usual today == climate change.
If you attribute every daily variance in the weather to the climate crisis, is it any wonder people stop believing or caring?