Stories
Slash Boxes
Comments

SoylentNews is people

posted by janrinok on Friday September 06 2019, @06:38PM   Printer-friendly
from the Ajit-Strikes-Again dept.

In May 2019, Neil Jacobs, the acting head of the National Oceanic and Atmospheric Administration (NOAA), testified before Capitol Hill that 5G wireless signals could decrease forecasting accuracy by 30 percent.

"This would degrade the forecast skill by up to 30%. If you look back in time to see when our forecast skill was roughly 30% less than it was today, it's somewhere around 1980," Jacobs said in May. "This would result in the reduction of hurricane track[ing] forecasts' lead time by roughly two to three days." A delay of two to three days could have a catastrophic effect on human life.

Still, these warnings haven't swayed regulators nor the cell phone industry. In August, Sprint announced more cities would be added to its 5G rollout plan. AT&T already has 5G available to corporate customers in various cities. Verizon already offers 5G to customers and has plans to expand, too.

"Right now the uncertainty is to what extent there will be an interference," he added. "In some sense the cause for education is to make sure that the existing weather sensing bands are protected and that 5G is in areas that are far enough away from where present weather sensors exist."

This does not mean 5G can't exist in states like Florida, but that the power might have to be turned down.

"If the power is turned down, there is a lesser likelihood that water satellites (that will sense the atmosphere) will sense the 5G network" instead, Gerth said.

Federal Communications Commission Chairman Ajit Pai has been dismissive of these concerns, which are only one of several in regards to 5G. As several experts told Salon last year, the effects of widespread use of mobile 5G need to be better-studied before it goes mainstream.

Why study when you can profit instead?


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 1, Informative) by Anonymous Coward on Friday September 06 2019, @11:28PM (9 children)

    by Anonymous Coward on Friday September 06 2019, @11:28PM (#890756)

    I know you're joking and I appreciate the humor in your comment.

    Drawing on an NHC forecast graphic with a sharpie was really stupid. The 120 hour NHC forecast cone never extended into Alabama except maybe overlapping a tiny corner of the southeast part of the state in one forecast. Doubling down on earlier statements in that way was a really stupid thing for Trump to do.

    However, there is a reasonable interpretation to some of Trump's comments. Although the NHC forecasts only go out 120 hours, I could envision Trump asking his advisors if there's any indication on what might happen beyond 120 hours. Most of the models produce forecasts well beyond 120 hours, even though the skill drops off. It's not unreasonable for a president to ask that question.

    It's entirely possible that someone showed Trump spaghetti plots of tracks through Alabama and Georgia. In fact, here are two spaghetti plots, one from 12Z on August 27 [ucar.edu] and 12Z on August 28 [ucar.edu] showing a lot of tracks from the GFS ensemble through those states. The GFS ensemble was farther west than a lot of the other models, as you can see with other tracks overlaid from 12Z on August 27 [ucar.edu] and 12Z on August 28 [ucar.edu]. In those last two links, the red lines are the GFS ensemble. AVNO is the "deterministic" GFS, AEMN is the average track from the GFS ensemble, and the other red tracks represent the individual GFS ensemble members. Those names are used because the GFS used to be called the Aviation (or AVN) model, and the spaghetti plots still use the older name of the model. The spaghetti plots do essentially represent a cone of uncertainty in the forecast. The GFS ensemble isn't that great of an ensemble because it's underdispersive, meaning that it doesn't capture the actual uncertainty in the forecast, but it is one of the ensembles that forecasters look at.

    If Trump's advisors showed him some of the spaghetti plots from the GFS ensemble that went beyond 120 hours, he might have really thought that Dorian was expected to go through Alabama and Georgia. A lot of the terminology has been conflated. For example, I saw an NBC News article [nbcnews.com] referring to the official NHC forecast graphic as a forecast model, which it isn't. "Model" refers to a computer model, whereas that's a human-generated forecast. With the terminology being conflated, it's plausible that Trump could have interpreted the ensemble as a cone and believed that Alabama and Georgia were within the cone of uncertainty.

    Drawing on the map with a sharpie was incredibly stupid. But it's completely plausible that Trump asked his advisors about where Dorian might go after 120 hours, the advisors showed a spaghetti plot of the GFS ensemble, and Trump believed that Dorian was most likely going to track through Alabama and Georgia. Considering how much the terminology gets conflated, I can see why Trump might have believed forecasters were expecting Dorian to hit those states. I suspect that Trump was hardly the only person who would have looked at early spaghetti plots and expected Dorian to make landfall along the Gulf Coast.

    I know, your comment is a joke, but I wanted to post this somewhere because there is a plausible explanation for some of what Trump has said about this.

    Starting Score:    0  points
    Moderation   +1  
       Flamebait=1, Redundant=2, Insightful=1, Interesting=1, Informative=3, Overrated=1, Total=9
    Extra 'Informative' Modifier   0  

    Total Score:   1  
  • (Score: 0, Insightful) by Anonymous Coward on Saturday September 07 2019, @12:21AM (1 child)

    by Anonymous Coward on Saturday September 07 2019, @12:21AM (#890778)

    In other words, he's too stupid to understand what he's been told, or more likely won't listen to what the experts are telling him about it because he doesn't have the capacity to understand it. Or he listened to an advisor (and almost certainly not a subject matter expert) who told him Alabama might be a possibility and "we did good there...."

    Any attempt at justifying that is simply making excuses for Trump's stupidity or failures as a leader. I don't care how he justified it. And I care far more about his not being able to admit he's wrong. That has a much higher capacity for being deadly for a whole lot more people than just himself or his staff whether that's civilians escaping a hurricane or soldiers he sends into harms way based on fucked up data he can't admit he misunderstood.

    • (Score: 0) by Anonymous Coward on Saturday September 07 2019, @02:17PM

      by Anonymous Coward on Saturday September 07 2019, @02:17PM (#890975)

      He probably mis-understood the XTRP model.

  • (Score: 4, Insightful) by ilPapa on Saturday September 07 2019, @01:13AM (6 children)

    by ilPapa (2366) on Saturday September 07 2019, @01:13AM (#890790) Journal

    However, there is a reasonable interpretation to some of Trump's comments.

    No, no, and no. If you look at the prediction maps for August 31, you will see that NONE (as in, not a single one) of the models had any lines even coming close to the Alabama border. Trump sent his tweet out on September 1, by which time it had been almost 36 hours since any prediction map had anything even moving in Alabama's direction. He claimed to be getting "hourly updates" while golfing. If he was actually getting "hourly updates", he would have known for over a day that Alabama was not in any danger at all. And those "120 hour maps"? They come out every few hours. There had been like eight superseding prediction maps AFTER the one Trump famously altered, and none of them showed the slightest possibility of hitting Alabama.

    So don't even try to suggest that there is any factual basis for any part of Trump's insanity. He says some shit that sounds good to him at the time, and then is 100% unable to admit to any mistake (which is not exactly a good trait to have in a leader). In this fat degenerate's case, he doubles, triples and then octuples down on the original lie because he just doesn't give a fuck and he's got the maturity of a 3 year old.

    I would say that altering an official prediction map (one with a footnote that says, "This map is superseded by any new maps") is just more scummy icing on the cake, except for the fact that he's still laying it on, considering the news that he's now ordered NOAA headquarters to claim that Trump was right and the actual fucking NWS meteorologists in Alabama were all wrong.

    --
    You are still welcome on my lawn.
    • (Score: 3, Insightful) by Anonymous Coward on Saturday September 07 2019, @06:01AM

      by Anonymous Coward on Saturday September 07 2019, @06:01AM (#890856)

      This is an example of why moderation of political discussions is awful. My post [soylentnews.org], which attempted to discuss the ensemble forecasts that were produced, is currently moderated -1 redundant. When it comes to discussing politics, moderation is essentially used as an agree/disagree vote, no matter how well thought out a post might be. It really irks me, as someone with a degree in meteorology, to put thought into some posts on this site, only to have some get modded to -1 because people have a hard time separating the politics from evaluating the quality of the post.

      Regarding the possibility of tropical storm force winds affecting Alabama, it's important to understand how to interpret the maps. The map that was doctored with a sharpie shows the expected track of the center of circulation and an estimate of the uncertainty in forecasting the location of the center of circulation. That map says nothing about the expected impacts in a given location. The size of the cone is not based on the spaghetti plots and the ensemble spread. In fact, it would be a pretty lousy idea to base the cone of uncertainty on that, because ensembles tend to be underdispersive. That means they don't account for the full range of uncertainty in a particular forecast. Instead, the cone is based on the error in forecasts of storm position from NHC forecasts in recent years. It varies from year to year, but not from one storm to another within a particular year.

      The extent of tropical storm force winds depends on the position of the tropical cyclone and the radius that the winds extend outward from the center. NHC produces probabilistic forecasts of tropical storm (34 knot), 50 knot, and hurricane (64 knot) winds at a particular location. Trump asserted that tropical storm force winds could affect Alabama. Here is the NHC archive of tropical storm force wind probabilities [noaa.gov] for Dorian. There are, indeed, graphics showing low probabilities (5-10%) of tropical storm winds affecting the southeast corner of Alabama for products issued as late as September 2.

      To be clear, Trump is being very dishonest here, particularly in doctoring the map with a sharpie. It's also ridiculous to request that NOAA release a statement supporting Trump's claims in order to benefit him politically. However, the NHC did, indeed, generate maps showing low probabilities of tropical storm force winds affecting Alabama. Advisory #26 shows the southeast corner of Alabama with 20-30% probabilities of tropical storm force winds. So it's actually false to say that NHC never produced a forecast showing the possibility for Alabama to be impacted by Dorian. For a sustained period of time, the graphics showed low probabilities across southeast Alabama. Trump is being dishonest, but the reporting on this issue isn't accurate, either. You can see the maps for yourself; that is a link to the National Hurricane Center, within the www.nhc.noaa.gov domain. Those are absolutely the official maps, completely undoctored by sharpies.

      The real problem here is politicizing matters of science, which generally leads to people being more confused about the actual science. We see it with climate change, where the scientists are nearly unanimous that humans are causing the Earth to warm. There is very strong observational and theoretical evidence supporting this, to the extent that it really shouldn't be in question at all. However, because of politics and misinformation, a significant segment of the population doesn't trust the science.

      This is an issue with high impact weather events, too, including a fake NOAA graphic that showed a 10 day forecast for Hurricane Irma to hit Houston [politifact.com]. Meteorologists are sometimes guilty of disseminating misleading forecasts, too. All too often, I've seen actual meteorologists tweet out outlier forecasts or forecasts showing high impact weather at some 240 hour GFS forecast when there's simply too much uncertainty to make such a prediction at that range. The public doesn't really understand probabilistic forecasts, so they don't really understand what a 30% chance of rain means. The real disservice is politicizing something that absolutely shouldn't be politicized. We don't need any more confusion about high impact weather forecasts than what already exists.

    • (Score: 2) by physicsmajor on Saturday September 07 2019, @06:38AM (4 children)

      by physicsmajor (1471) on Saturday September 07 2019, @06:38AM (#890862)

      None, not a single one of the models is what you stated. So a single example should be sufficient to disprove your entire premise, then?

      Let's see what CNN was showing: https://twitter.com/realDonaldTrump/status/1170089069105340416 [twitter.com]

      • (Score: 1, Interesting) by Anonymous Coward on Saturday September 07 2019, @01:21PM (3 children)

        by Anonymous Coward on Saturday September 07 2019, @01:21PM (#890953)

        I've pointed out in this thread that some ensemble forecasts from August 27 and 28 showed Alabama and Georgia at risk from Dorian. When Trump said that early on, some models showed Dorian hitting Alabama and Georgia, that's actually true. It's also true that very low probabilities for tropical storm force winds were maintained in NHC forecasts for far southeast Alabama into September 2. These are facts. I didn't check your link, but it's possible you might have found an outlier model from September 1 showing a track into Alabama. I'd point out that it's an outlier and a very unlikely scenario.

        While there are some facts that technically support Trump, he also made some very misleading statements later on about the risk to Alabama. Editing the NHC forecast with a marker is inexcusable. Rather than clarify his statements or admit he made a mistake, Trump has gone to absurd lengths to claim that he was correct. It seems likely that someone higher in the administration directed NOAA to criticize WFO Birmingham for their tweet. That's unacceptable.

        It's a fair point that the NHC graphics differed from the statement from WFO BMX, and that it might be confusing to the general public. I suspect the wind speed probabilities are based on typical NHC forecast errors from recent years, along with the forecasted radius of the winds in the tropical cyclone. This could be fixed by giving NHC forecasters the ability to edit the cone of uncertainty depending on the clustering of computer models and their confidence in the forecast track. If the model guidance was tightly clustered keeping Dorian over the Atlantic, NHC forecasters could coordinate with local weather forecast offices (WFOs; like Birmingham) to update the forecast and maintain a consistent message between local and national offices. This is exactly what's done for severe thunderstorm and tornado watches, where the Storm Prediction Center (who issues the watches) coordinates with the local WFOs about which counties should be included in a watch.

        I would like to see more consistent messaging between NHC and local WFOs. NHC makes general hurricane forecasts but also directs users to local WFOs for more specific guidance about the impacts expected at any particular location. However, the general public probably isn't aware of this, so the inconsistent messaging is an issue. The solution is to make the messaging more consistent, not to throw the WFO BMX forecasters under the bus. NWS forecasters are civil servants, not political appointees. They're tasked with protecting life and property and are assigned a county warning area (CWA) for which they issue forecasts. The job of the WFO forecasters is to convey the most accurate information to the people residing within their CWA, which is what they did. Trump isn't interested in improving the messaging to the general public. He's only interested in trying to vindicate himself, even if it means scapegoating civil servants. I don't work in the National Weather Service, but I'd be really pissed off right now if I did.

        For that matter, I'd like to see NHC issue 6 and 7 day forecasts for tropical cyclones, because the forecasting skill has improved quite a bit over the past couple of decades. They used to only forecast three days out, but eventually added 4 and 5 day forecasts. As skill has continued to improve, it would make sense to make official forecasts at longer lead times. People are going to look at the ensembles and they're going to be curious about the tracks beyond the current five day forecast period. In the absence of official information, people may get misleading or outright false information from sources like social media. Making official forecasts out seven days would probably limit the spread of misinformation.

        As for Trump, he's correct that some models did suggest a threat to Alabama early on, though pretty much everything else he's said is stupid and driven by an inability to admit he made a mistake. The NOAA statement is correct that low probabilities for tropical storm force winds were shown for parts of Alabama, though the rest of the statement is asinine. This is damaging to the credibility of forecasters and the National Weather Service. How can NOAA appear credible testifying about issues related to 5G when they issue bullshit statements like throwing WFO BMX under the bus? Will people be less willing to trust the forecasts from their local WFO after this nonsense? Politicizing weather alerts is very dangerous and this needs to stop.

        • (Score: 2) by legont on Saturday September 07 2019, @05:58PM (2 children)

          by legont (4179) on Saturday September 07 2019, @05:58PM (#891035)

          Like I mentioned many times before, the probability does not mean anything. What matters is the probability multiplied by the damage.

          In this sense, a prediction that a hurricane hits Florida does not mean much. Florida folks supposedly know about the danger and are prepared for it. They will simply collect insurance and if they are not insured - too bad. Should never leave NY to begin with. Similar are Bahamas that do not deserve the US rebuilding help. They should take on the costs. Why should I pay for their fun in the sun?

          Back to Alabama, it is less probable path and damages would be way more significant. It is not reasonable for Alabama folks to be ready for a hurricane. Therefore, they should be warned and if it does hit, helped.

          --
          "Wealth is the relentless enemy of understanding" - John Kenneth Galbraith.
          • (Score: 2) by ilPapa on Saturday September 07 2019, @06:33PM

            by ilPapa (2366) on Saturday September 07 2019, @06:33PM (#891049) Journal

            Therefore, they should be warned and if it does hit, helped.

            But they should NOT be warned four days after the risk of being hit becomes zero, which is exactly what the big soggy baby in the White House did.

            Remember, the exact terminology Trump used was Alabama "will most likely be hit (much) harder than anticipated". Every single word in that phrase is absolutely wrong.

            --
            You are still welcome on my lawn.
          • (Score: 0) by Anonymous Coward on Sunday September 08 2019, @04:54AM

            by Anonymous Coward on Sunday September 08 2019, @04:54AM (#891188)

            Just to be clear, we're talking about tropical storm force winds here, as in 34 knots or 39 mph. If you get 39 mph winds, it's a windy day, but it's not going to cause much damage. The potential for damage from 39 mph winds is quite low. We're talking about a low probability multiplied by low damage.

            NHC also produces maps for 50 knot (58 mph) winds, and here's the archive of those graphics [noaa.gov] for Dorian. Advisory #25 has about half a county within the 5-10% range. Advisory #26 has several counties within the 5-10% range, but nothing higher. To be clear, 58 mph is also the minimum threshold for the National Weather Service to issue a severe thunderstorm warning based on damaging straight line winds. At 58 mph, the winds can cause minor damage, but a low end severe thunderstorm warning is not going to do widespread damage.

            We also refer to 65+ knot (75 mph) winds as a threshold for significant severe weather, meaning that there's the potential for higher end damage when the winds reach this strength. That's also essentially the threshold for hurricane force winds (64 knots, 74 mph, but we're splitting hairs here). When severe thunderstorm warnings list the expected hazards, the wind hazard is given in 10 mph increments, so 60, 70, 80, 90, and in rare cases 100 or 110 mph. Typically you'll start to see enhanced wording in the warnings like "THESE ARE VERY DANGEROUS STORMS" once the winds get to 80 mph. In other words, higher end damage is most likely when the winds reach hurricane force. NHC also has graphics for hurricane force wind probabilities [noaa.gov] and the 5% contour never entered Alabama.

            Back on August 27 and 28, some models did forecast Dorian to cross Florida and then make landfall along the Gulf Coast. Forecast uncertainty was much larger at that time, but there were plausible scenarios where there would have been a substantial risk of damage in Alabama. After that, models converged on a solution in which Dorian would recurve without crossing the Florida peninsula. There was still a legitimate possibility for landfall in eastern Florida, but by September 1, only a couple of outlier solutions in any of the ensembles would have resulted in any significant threat to Alabama. The ECMWF ensemble [ecmwf.int] didn't suggest any significant threat to Alabama by September 1. Neither did the GFS ensemble or the Canadian model, as you can see from the 00Z September 1 [ucar.edu] graphic. Only a couple of outlier solutions from the NAVGEM ensemble suggested any significant risk to Alabama.

            Before you say, what about the NAVGEM ensemble, it's not that great of a model. Although this slideshow [wmo.int] is three years old, it has a lot of great information that isn't particularly outdated. The NAVGEM model configuration is inferior to most of the other models that are used in forecasting tropical cyclones. The NAVGEM also didn't perform well in the verification of track forecasts. If the only model solutions with any significant impact to Alabama are a couple of outlier solutions from the ensemble of a model that is known to be inferior, you can pretty much rule out that scenario.

            There are really only two things that Trump has said about Dorian and Alabama that are in any way defensible:

            in fact, under certain original scenarios, it was in fact correct that Alabama could have received some ‘hurt.’ Always good to be prepared!

            This was on September 2, and it's true that some of the older ensembles from around August 27 and 28 did suggest a plausible risk to Alabama.

            Later, Trump tweeted a South Florida Water Management District map dated from Aug. 28 that showed that some projections showed that the hurricane could reach Alabama. The graphic notes that advisories from the National Hurricane Center should “supersede” the map.

            This quote is from an article in Time [time.com] and attributes that to a tweet on September 4. While the map from August 28 is correct, it was also a full week out of date by that time. Again, there was a plausible threat to Alabama back on August 27 and 28.

            Trump was correct that some early scenarios back on August 27 and 28 did indicate a legitimate risk to Alabama. No problem. The map he tweeted on September 4 contains accurate information, but was a full week out of date by then. The rest of what Trump said is indefensible. When he tweeted on September 1 that the risk to Alabama was much greater than previously forecast, that's outright false. In fact, the risk to Alabama had become negligible by that point as models had trended eastward with Dorian's track.

            If I was rating this for Politifact, I'd rate one of Trump's statements as true, one as half true (because it's misleading, with out of date information), and the remainder as false or pants on fire. I'd rate your statement about the risk to Alabama as false, for the reasons I just outlined.