...okay so what actually happened, as I understand it

some researchers looked at the reported numbers of people displaced by rural dam reservoir projects, and found that the all standard grid-based population density models tend to underestimate rural populations--not that they're all offset consistently by some amount, but that their mean is offset, with a huge scatter

and they consistently underestimate more often than they overestimate, relative to the reported numbers

...but they also acknowledge that their data is just... whatever was reported by various dam building authorities around the world, and they don't have access to the methods of data gathering

They state it seems unlikely that anyone would deliberately exaggerate the number of people displaced by a dam project, but they also don't have any proof that those numbers are especially reliable, so it's basically a "they said X and I can't think of a reason it would be less than that" argument

their point being "hey we used this data that's lying around and it disagrees a lot with those other data we've been using, so we should probably check to make sure our standard data sets are good," and not "there are billions more people than we thought!!!!!!!!!!"

especially since, as they say themselves, the standard data sets are way better for urban and suburban populations

I'm a little puzzled exactly how they did their analysis, despite reading their Methods section, since in the one example they show in detail, it.... sure looks like the population density model is pretty accurate to their relocated population number

but it's also not clear what the numbers actually represent or why--as far as I know the International Commission of Large Dams doesn't have the ability to reward or punish based on the accuracy of reported numbers, so who knows where the administrators handing them those numbers got them from

...it seems they also dig up an estimate for the average time to complete a dam, and used that for the offset between date of completion and the population density year, which... is reasonable, but i can think of lots of ways for that to return profoundly flawed data when applied to hundreds of different projects of different scales

that's hiding a lot of idiosyncrasy and complexity behind a mean

Yes.

But I also would cheerfully believe that the models for calculating world population underestimate rural areas dramatically because there is still a lot of if not downright regressive reluctant to change among farmers and so the idea that lots of children is still

Economically beneficial ++ farmers do have the most sex (journalists have the least lmfao) and there are certain discrepancies in things like Afghanistan and Africa, and Pakistan where I have almost 0 doubt that they just aren't counting people properly

it's definitely worth checking

i mean they talk about some newer models that apparently can consistently pick out individual and clusters of buildings in the landscape but don't agree with their people-reported-displaced-by-dams metric

which... that seems like it would be a pretty sound measurement, the way they describe it???

my point is just that it's an interesting independent measurement inconsistent with the accepted model... but it's also just an interesting independent measurement inconsistent with the accepted model

those happen all the time, and the vest majority are due to errors or factors not accounted for in the analysis

*vast

so the whole breathless science journalism thing of "OOPS ADURR HURR SCIENTISTS GOT IT WRONG AGAIN BETTER START COUNTING FROM 1 LULZ" is like 98% empty hype

Maybe. I'd be curious if we wound up at like 11 billion and just 3rd world infrastructure was that extremely inadequate etc.

rofl I mean, I can't say it's impossible

I feel like the real takeaway is "man our population data kind of sucks, we really need to figure out a way to make it more robust"

Yes!!! Agreed lol

imagine if science journalism actually reported what's actually being said and done instead of just OMG EVERYTHING WE KNEW IS WRONG clickbait horseshit

imagine. Let's make it happen