Similar topic was discussed earlier: https://news.ycombinator.com/item?id=44409175 (140 comments)
Loss of key US satellite data could send hurricane forecasting back 'decades'
51–60 of 167 posts
Re: Loss of key US satellite data could send hurricane forecasting back 'decades'
#52Earlier quoted context omitted.
> To anyone acting as if this is a surprise or they're suddenly caught out and have to switch to another provider, I have to wonder, with the writing on the wall for 8 years now, how have you not already updated your plans? That doesn't accurately capture the reason why there's outrage here. In the weather community, we're constantly thinking through contingencies because a great deal of things are out of our control…
> Some private companies operate PMW satellite constellations and we could accelerate the acquisition of these data, but there are limited (read: none) federal mechanisms to do this and due to vertical integration in the weather industry, the operators of these constellations may not actually be inclined to do so - and certainly won't do so on the cheap, especially for the federal government. That's the goal, actuall…
The companies I'm referring to are (generally) not publicly traded, so it's not quite that simple. Is it possible that some sort of backroom shenanigans are going on here? Yeah, absolutely, especially as several knowledgable folks speaking publicly about this episode are pointing their fingers at opaque procedure within Space Force.
But Hanlon's razor ought to apply until shown otherwise.
Re: Loss of key US satellite data could send hurricane forecasting back 'decades'
#53Earlier quoted context omitted.
They could claim that even with the satellites. The "alternate reality" can be anything - if facts aren't inserted into it the people inside won't know.
Idiots will buy it. The courts won’t. Cutting off the data stymies the latter.
The courts are compelled to defer to SCOTUS, which has demonstrated that it is ideologically aligned with the regime.
Re: Loss of key US satellite data could send hurricane forecasting back 'decades'
#54Earlier quoted context omitted.
That's not at all how insurance companies price risk. Unknown risk is more risk, and more risk is more expensive. Therefore, unknown hurricane data is more risky and thus more expensive. If you know your car's engine is going to need replaced after exactly 100,000 miles, you know to save up for a new engine or a new car - and you know how long you have to save, so you can precisely set aside an appropriate figure eve…
Hurricane risk has been grossly exaggerated for years. Every year people say it will be the end of Florida as we know it. But those promised hurricanes never come. The worst is some flooding and damage at coastal areas, but it’s always anti-climactic. The real reason insurance is high is because of fraudulent claim risk. Hurricanes themselves are more or less a solved problem in Florida. That data is useless.
> Hurricane risk has been grossly exaggerated for years
Year-over-year, economic impacts and disruptions due to tropical cyclones are dramatically rising. Most of this is an exposure issue. But long-tail events - like Andrew's utter devastation of Homestead in 1992 or Katrina's unique confluence of storm surge in urban/suburban parishes in LA - can and do happen.
One day, there will be another Galveston or Homestead.
Re: Loss of key US satellite data could send hurricane forecasting back 'decades'
#55“Researchers say the satellites themselves are operating normally and do not appear to have suffered any errors that would physically prevent the data from continuing to be collected and distributed, so the abrupt data halt might have been an intentional decision.” Wait, the U.S. aren’t even going to try selling the satellites? We’re just scrapping them?
That claim does not seem justified. > 2016 failure of DMSP 19 without replacement[edit] On 11 February 2016, a power failure left both the command-and-control subsystem and its backup without the ability to reach the satellite's processor, according to the U.S. Air Force Space Command investigation released in July 2016 that also announced that DMSP 5D-3/F19 was considered to be 'lost'. The satellite's data can still…
Ah, so basically if you have a car that's 5 years out of warranty but still runs fine, and the government comes in and takes your keys so you can't drive it, that would be your fault for not having gotten a new car sooner?
Re: Loss of key US satellite data could send hurricane forecasting back 'decades'
#56Earlier quoted context omitted.
Did DOGE not ditch hundreds of probationary employees at NOAA, cancel numerous contracts, get 1000 people to take early retirement offers, get rid of buildings, etc? And now the current funding request enacts a ~30% funding cut. I'm not sure the factual issue you're seeing. Is it that the statement wasn't definitive enough in saying that DOGE apparently was a large part of instituting these cuts? (Yes, I know OPM imp…
My understanding is this was set up to happen roughly a decade ago and is just now manifesting. It has pretty much nothing to do with DOGE.
We don't if, or to what extent, DOGE was involved or influential in the decision-making here.
Yes, the DMSP program was aging and slated to wind down as replacements - both federal and commercial - came online in the second half of the 2020's. But in general, if valid and useful data continues to stream from these types of satellites, you use it and monitor for disruption.
As someone who uses the DMSP data every single day, let me be very clear: there was no warning or expectation that such an abrupt change was going to happen. Yes, we all have contingency plans for if a satellite fails or a data link goes down. But to be given basically 5 days notice that a significant, mission-critical asset would be taken offline? That doesn't - and shouldn't - happen.
Re: Loss of key US satellite data could send hurricane forecasting back 'decades'
#57The problem of important projects surviving political change is a tough one.
A lot of these important projects have a single point of failure - who is the president every four years. I wonder how we build institutions and resources resilient to that?
I realize privatization is an ugly word, but could some of this stuff be provided by the private sector?
Can we make it possible to fund initiatives in a multinational manner where countries contribute to these efforts, but if one country blinks out, then you still have it go along?
Re: Loss of key US satellite data could send hurricane forecasting back 'decades'
#58Earlier quoted context omitted.
My understanding is this was set up to happen roughly a decade ago and is just now manifesting. It has pretty much nothing to do with DOGE.
We don't if, or to what extent, DOGE was involved or influential in the decision-making here. Yes, the DMSP program was aging and slated to wind down as replacements - both federal and commercial - came online in the second half of the 2020's. But in general, if valid and useful data continues to stream from these types of satellites, you use it and monitor for disruption. As someone who uses the DMSP data every sing…
[flagged]
Re: Loss of key US satellite data could send hurricane forecasting back 'decades'
#59Earlier quoted context omitted.
That's not at all how insurance companies price risk. Unknown risk is more risk, and more risk is more expensive. Therefore, unknown hurricane data is more risky and thus more expensive. If you know your car's engine is going to need replaced after exactly 100,000 miles, you know to save up for a new engine or a new car - and you know how long you have to save, so you can precisely set aside an appropriate figure eve…
Hurricane risk has been grossly exaggerated for years. Every year people say it will be the end of Florida as we know it. But those promised hurricanes never come. The worst is some flooding and damage at coastal areas, but it’s always anti-climactic. The real reason insurance is high is because of fraudulent claim risk. Hurricanes themselves are more or less a solved problem in Florida. That data is useless.
> Hurricanes themselves are more or less a solved problem in Florida.
I'm going to go with less, though I suppose you could call "experience widespread destruction, get bailed out by the federal government, rebuild in the same spot" to be a permanent solution.
Florida has maybe solved cat 1-2 hurricanes.
Re: Loss of key US satellite data could send hurricane forecasting back 'decades'
#60Earlier quoted context omitted.
> We can't actually say this was DOGE The article is saying it was DOGE. DOGE directly attacked our hurricane-forecasting capacity [1]. OMB, i.e. Vought, continues that attack [2]. Given the top three states by hurricane risk voted for Trump in ‘24 [3][4] this should make for an entertaining hurricane season. (Particularly if both a red and blue state get hit and request federal assistance.) [1] https://apnews.com/ar…
[flagged]
> "Entertaining" in that red states will get their requests approved
Between the cost of the damage (and us blowing the card preëmptively on this big, beautiful bill), reduction in state capacity to respond to disasters as a result of DOGE and an increasingly-fracturing Congressional GOP I'm not sure they will.