I believe once they are commissioned they are given a number (eg current ones are GOES-13, GOES-14, GOES-15). They are also given a designation as GOES-EAST or GOES-WEST depending on where they are being used. So GOES-13 was GOES-EAST until it had an anomaly and GOES-14 was moved out of a storage orbit and designated GOES-EAST. It returned to storage once they fixed the issue with GOES-13.
The R is just a sequence number for launching; S and T are slated next. If there is a failure before commissioning, it won't receive a numerical designation.
So the GOES-R name was just a happy accident I guess.
“This is probably the most complex technology that our species is involved in — weather prediction is,”
I think this is just a pattern of people seeing the complexity in their field, but not knowing anything about other fields and thus believing that they are doing the most awesome thing ever.
LHC guys will say their technology is the most complex one, Intel, the AI guys, and so on..., not to mention the global infrastructure required to run the Internet.
No, they are using complex in the it's got may interlocking parts and all those parts are fiendishly difficult to understand individually and put together sense :)
They're referring to the total technological package, supercomputer forecasting, in addition to satellite building, lofting, and monitoring.
It's not a reference to the complexity of chaos theory. At least that's what I get from the context of the quote, which continues, “And we do it together as one species. It’s one of our greatest successes.” Made me get the warm fuzzies, sometimes we're not so shabby. :)
That definition of complexity, counting all the interlocking parts, leads to laughable conclusions. Example: The decision whether to walk my dog at the park or at the beach is now more complex than launching satellites. Since the park/beach decision relies upon both weather forecasts and dog psychology it is therefore more complex than either. Since weather forecasting involves launching satellites, there are more parts in the park/beach decision.
To be relevant, complexity should measure not the entire system but the number of variables at a specific point in the system. Rocket science isn't very complex as it is very compartmentalized. Different stages in the process involve discrete decisions divorced from others scheduled to occur at different times. So, weather forecasting wins because the number of variables at every decision is far greater. The flow-chart for a rocketry decision is more vertical whereas the chart for a weather forecast is more horizontal.
After reading Nate Silvers book about predictions where he delved into different fields I believe this is pretty much correct if it had said the most complex predictive science. The other one that might be harder is earthquakes since we haven't even gotten close to guessing when those are coming.
In terms of raw computing power I don't think there's any industry that utilizes as much super computing time as weather prediction. So there's also that.
Weather forecasting for Norway keeps about 500 dual-socket Xeon E5-2630 servers busy 24/7 all year. Norway has 0.3% of the world's land area, so a rough estimate, assuming Norway uses 5x more CPU-hours per sq. km. than the world average, would put global weather prediction use at about 30 000 nodes running 24/7.
For comparison, a dedicated AWS m4.4xlarge instance, which is equivalent to one of these nodes without the expensive Infiniband, is about $500/month. For an actual supercomputer node you can figure about double that.
So the total running cost (of just the hardware) for global weather prediction per month is probably around 30 million USD. Then you also have to pay people to manage the running of simulations, do the analysis, produce nice plots and animations for TV, etc.
Can you scale the prediction effort like this? Don't they run a global model with just a focus on the "home area", or are the resolution differences so large that the "rest of the world" just isn't very expensive to calculate? Or are there shared global models that are used as input data?
Those are very good questions. My post above attempts napkin math, so it could probably be off by a factor of 10, but probably not by a factor of 100.
Also, I Am Not A Weather Forecast Scientist, just working in a related field (fluid dynamics). So I could be wrong about stuff here.
As for shared models and data: I know Norway and the UK do a lot of this. I imagine so does the US and Canada/Mexico.
Also, keep in mind that weather modelling relies heavily on running an ensemble with random variations in initial/boundary conditions. So a lot of the computational cost comes from running hundreds of these simulations. My guess is that you don't run the large area 100 times, just a few times and then perturb it for all your local area simulations.
Fun question, even if it wasn't a serious one. Since I happen to work at a company whose business is related to ad placement prediction, I can try to answer.
In our case, we have 1000-2000 instances (the number is elastic, because AWS) handling real-time ad auctions and related data processing. Since we are just one of the many players in the industry, and far smaller than the largest ones (GOOG, FB), I am pretty sure in aggregate more computing cycles are used to optimize ads than oil or weather.
The oil industry uses quite a bit of computing power, mostly to process seismic data. Here's some "sponsored content" (found via a quick Google search) about BP's computing center in Houston: [1]
Some statistics extracted from the article (written by a non-technical journalist, I suspect, so treat them accordingly): 5,200 servers, 115,000 CPUs, 3.8 petaflops, 27 petabytes.
Bear in mind that BP is only one oil company, and by no means the largest (you probably think Exxon is the largest, but Saudi Aramco is much bigger [2]).
The global internet infrastructure isn't the most complex, but it might win for the most jury rigged. Everyone in technology (I'm a Linux Systems Admin / Software Engineer) knows that the internet was built by a team of macguyvers and is held together by shoe string, paper clips, duct tape, and undersea cables.
I've worked with particle accelerators, cancer data, and AI and i'd agree that weather prediction is probably the most complex technology we have at the moment.
Weather is both easy to measure and quite hard to predict and we've studied it for longer then most things. So we have a vast global array of sensors and ensembles of complicated models running constantly on supercomputers.
I hope we get to that level with biology and other fields at some point but we aren't there yet.
I feel for meteorologists. They absolutely have the rest of the world beat on the discrepancy between modeling complexity and ease of observable result.
"Well, they said it would rain and it didn't..."
So, from me at least, thanks for all the hard work!
Maybe one of the problems is communication. The app Darksky seems to do a fine job of predicting rain within a short time window because it knows my location. The odds of old school broadcast communication meteorology making a reliable predictions outside of generalities seem slim to none.
Capital Weather Gang usually does a better job explaining things to the masses, so that's a little disappointing. Having a similar conversation with family and friends, it's challenging to explain the hype.
Weather is incredibly complex and difficult to model and predict. We have multiple sources of truth to provide the input to our models: radar sites spaced hundreds of miles apart, balloons that are launched twice a day and hundreds of miles apart, METAR stations that record ground observations, satellite, etc. Satellite data is the closest you can get to seeing the entire atmosphere as an interconnected system with no gaps in data, but is limited by being a top-down view (better spectral information will help), somewhat low resolution, and 5-15 minute update intervals (storms can go from nothing to tornadic in 45 minutes).
CIMSS and other groups of researchers are doing great work with our current satellite data like identifying when supercells are likely to produce tornadoes and estimating instability in a column of the atmosphere. Increasing spectral information, resolution, and frequency will enable products to be refined and new products to come out - such as lightning detection, which currently relies on private networks.
Severe weather nowcasting is only one of the areas that will benefit from this, but the one that I'm familiar with as an amateur meteorologist. GOES-R has the potential to cut false alarm rate and increase warning times for severe weather. Weather nerds are rather excited, as this stands to be a powerful new tool.
GOES-R represents a nearly 2 orders of magnitude improvement over previous generations. Imagery will have higher spectral and spatial resolution, by significant margins, and provide coverage in new areas. Consider just one aspect: lightning observation, GOES-R will vastly increase the amount of data available. This is a huge amount of data that wasn't previously available and will no doubt be able to be used in improving weather modelling. However, lightning intensity has been recently discovered to be a key predictor in thunderstorm intensity and tornado formation so having that data available in real time will likely end up leading to better tornado predictions and many lives saved. And that's just one element, every aspect of the satellite is a huge leap over previous generations and will have a tremendous impact on weather forecasting and climate research for years to come.
When will humanity have enough sensors and computing power to give a similar forecast that Dark Sky provides today, but 7 days in advance? What will that system look like? It seems like it is a mostly matter of time, but I just don't grasp what all will need to be done to get there and if it is 15 years out or 100.
The only real hope for that is if we start directly controlling weather. The last 1,000,000x increase in processing power has bought less than a days worth of accuracy. In the very short term you can just model the momentum of the air, but the feedback loops quickly reduce stability.
This is actually pretty cool. I last worked with satellite data from GOES 8 and 9 in '98. I remember satellite images were sent sent every 3 hours and full disk images were 12 hours. I worked in a weather office where they had one pc running DOS that received the data from Satellite and saved the data where various HP-UX servers manipulated the data.
Last i remember, all data transmitted by these satellites are not encrypted, anyone can build a receiver to gather the data to generate satellite images.
Isn't there a lot of work that goes into post processing the data? It wouldn't matter if you just want an awesome background image, but using the data for some kind of actual work... there is a lot of calibration to be done. I was working with some ESA Sentinel data several months ago - not fun.
Edit: For those who don't normally work with this stuff, I'm talking about not only correcting geospacial data - but also bringing the sensor data in line with a bunch of other sensors and past datasets. It quickly goes from "Oh, pretty pictures!" to "Do I really want to do this anymore? Can I ever trust data again?" That said, doing your own ground segment is cool - just know that there is a whole lot more work to do beyond actually receiving the radio transmissions.
This is great. I worked on the ground processing systems for this satellite. I left before my own algorithmic systems went through final review. Was good for me to leave career-wise but I wish I could have seen it through to the end.
Woah, something from the next generation? Someone call the temporal agents and let them know. This is exactly what we faught against in the temporal cold war.
Enjoy it while you can, as most likely the incoming administration will turn off this satellite. Why spend a lot of money to discover politically inconvenient data? Two birds, one stone.
The parent is downvoted but raises the noteworthy point that congress has in fact diverted funds away from atmospheric research specifically to avoid climate science. I'm not sure where this GOES satellite falls in the constellation of climate study tools.
Ironically, they can't fully defund it as it is mission critical for the military industrial machine. All deployed aviation assets in the US military have somewhere attached to them a USAF meteorologist. I flew the Shadow 200 TUAV "drone" in Iraq from 2003-2004 and we always had a single air force person with us for weather stuff. It is critical for safety of flight. Also, for launching things like missiles and really anything airborne, you need to know the weather for safety reasons. It only lightly touches on it in the article, but this is why the DoD is so serious about funding and working with NOAA even though NOAA is part of the Department of Commerce.
> they can't fully defund it as it is mission critical for the military industrial machine
The U.S. military lacks critical funds at this point due to defunding, depending on how you define "critical". The service chiefs testified that they will not be able to defend the country unless they get more money.
They've been saying that since the dawn of the US Military. I genuinely don't buy it. It is just like Trump's alarmism over the size of our military (the navy specifically) shrinking under Obama post-GW Bush surge. The technology has gotten so much better and automation in specific allows us to do more with less people. This is by design and I'd almost argue if the tech wasn't getting better and needing less humans, our R&D is clearly failing. The current Order of Battle of the entire US Military calls for us to be able to fight two large scale wars simultaneously. We've done that arguably well/poorly with Iraq and Afghanistan. They're coming to grips that with unconventional warfare, you simply need a LOT more boots on the ground trudging around.
Our technology advantage alone allows us to realistically probably coast for 6-8 years and still be matched only realistically by China. Russia has a few things that can match current US kit (their new tanks are excellent and they have some of the best anti-air kit in the world), but the Kremlin simply far and above lacks the cash to be in a position to afford to pay for a serious modernization. Obama's sanctions along with the massive drop in oil have put a dent in Russia's economy regardless of Putin's posturing. Sure they have very nice new jets, but they have precious few of them. We have nice new jets, better radars, and all of our older jets can spank the pants off of their older jets, etc, etc.
This is really just my anecdotal experience as a US Army veteran who was in OIF II 2003-2004.
> They've been saying that since the dawn of the US Military.
I follow foreign policy issues closely, and I disagree.
> The current Order of Battle of the entire US Military calls for us to be able to fight two large scale wars simultaneously. We've done that arguably well/poorly with Iraq and Afghanistan.
I'm not sure what you mean by that usage of "Order of Battle". IIRC US military planning calls for them to fight one war, a medium-sized one IIRC, and keep a second medium-sized war in a holding pattern. Iraq and Afghanistan are small wars against very weak opposition whose best weapons are IEDs.
However, there isn't much here beyond claims with no basis. I'd be interested in seeing some evidence for it all.
Having been in the US military, the command was always alarmist about troop sizing even in the blatant face of improving technology. We have counter-defilade rifles now (XM25). This puts the best part of a small mortar in the hands of an infantry rifle platoons. Does it mean you can get away with less artillery platoons and mortar squads? It might, yes. Does it mean you can retrofit counter-defilade capabilities into normal infantry operations allowing smaller units to do more? It does. This was my point. I saw this directly while I was in the service and it has always been that way. Yes we've done budget cuts under Obama for the US military, but we also had an unprecedented "surge" during GW Bush, meanwhile tech has been improving. Look at an older Nimintz class Aircraft Carrier with their 5000+ crew. Compare that to the newer Gerald Ford class carriers with their increased operational velocity and 2600 crew. Will some admirals tell you they're hamstrung due to having fewer soldiers? No doubt. Does that always reflect reality since they can literally do more with less? Nope!
The "Order of Battle" is US Military terminology to denote the composition of a military unit to respond to certain threats. It is most commonly used from Division / Brigade level and lower by the intelligence officers, but goes all the way up to the Pentagon & Department of Defense. An order of battle is designed to solve the question of, "How do I shape a military unit to best defeat the adversaries they're pledged or expected to fight against?". Absolutely nothing in the US's two war doctrine says for us to fight a medium war and be prepared for another medium sized war. Not sure what you think, but Iraq and Afghanistan are both big wars where we've simply de-emphasized "winning" them to optimize for less soldiers dying. The land mass is enormous in both and since a war is basically about owning a piece of land, it is easier to judge the size by the size of the land being attempted to liberate. Also, if we were fighting a larger conventional war (sans nukes) we would likely win/lose much quicker with much lower casualties. The asymetric warfare we're currently waging is something the US simply isn't very good at (nor is any conventional military really). What do you think?
For instance, I was a member of the US's first "Stryker Brigade Combat Team" (D Troop, 1/14th Cavalry, 2nd Brigade, 2nd Infantry Division out of Ft Lewis, Washington). Our Brigade of ~5500 soldiers replaced the 82rd Airborne division of ~19000 soldiers and did equally well due to our Order of Battle emphasizing speed (via the Strykers mainly) over sheer massive footprint and heavy slow armor. This was due to the enemy not being heavily armored and primarily using small pickup trucks "technicals". In this area, we pretty much dominated it as you'd have a fully loaded Stryker squad across the city (Tall'Afar or Mosul primarily) before anyone even realized it as they were much quieter than say a Hummvee.
I think people are accustomed to it and therefore overlook it, but note that government, using taxpayer dollars, delivers the most high-tech systems in the world - not only for meteorology, but also science, medicine and space exploration (as well as military and intelligence). There is a benefit to countries and to the world when the people pool their money for something big.
Government doesn't do shit, except write the check. this all comes from acadamia, nasa, darpa, etc.
You're point that science is pushed forward by people pooling resources is absolutely correct, I just want to make it clear that Government has very little to do with it. This is just societ bolstering the resources of our greatest minds.
In a mathematical sense, yes, every citizen who gives a certain amount of tax money to the government to pay for research bodies could, instead, give money straight to some sort of central body to distribute it to the same various groups.
But they wouldn't. And you know it. Even the Science Philanthropy Alliance (http://www.sciencephilanthropyalliance.org/what-we-do/resour...), which funds researchers using privately-donated money, says that taxes are necessary to have scientific progress.
I didn't say anything IF taxes were important or IF people's WOULD pay them.
And yes, we are talking about semantics (mathematical sense). People couldn't give money to the DoD and accomplish the same thing, nor could they give money to DHS or any of the other Executive Departments.
NASA is an independent, federally funded, civilian agency. The part of it that is Federal, is its budget.
Sort of, yes. Neither one of them are actually PART of the federal government. They are separate, independent organizations whose largest (and probably exclusive) customer is the United States government. NASA for example doesn't actually rollup to any cabinet position.
As opposed to the Department of Energy or Department of Education for example.
> Neither one of them are actually PART of the federal government.
Indeed they are. Their budgets are paid for by taxpayers and established by laws passed in Congress and signed by the President. Their employees are federal employees, AFAIK.
> NASA for example doesn't actually rollup to any cabinet position.
The agencies report to the President, whatever the org chart is. The President chooses their administrators, and directs their operations (assuming she/he has time and interest).
DAPRA is in the Department of Defense.
Finally, I believe the CIA was an independent agency, like NASA, before the DNI was created about a decade ago. Was the CIA not actually part of the U.S. federal government?
The NASA Administrator, Charlie Bolden, was nominated by President Obama and confirmed by the United States Senate. NASA is not part of any department but rather reports directly to the President of the United States.
I understand that they are PAID by the federal government. What I'm saying is that they operate INDEPENDENT of the government. They don't have any role in governance, they don't draft or pontificate on legislation, and they don't manage or participate in the daily administration of government affairs.
The ONLY thing Federal about them is their funding.
No. NASA and DARPA are managed, funded, directed, controlled, cleared and authorized by the Federal government. There's nothing not Federal about them. And they do not operate independent of the Federal government.
Perhaps you're getting mixed up on the nomenclature?
There are "independent agencies" in the US Executive Branch, but these aren't independent of the entire US Government; they're just not part of a single Department. For example, NASA could have been part of the Department of Transportation (and thus report to the Secretary of Transportation), but it isn't and reports directly to the President.
Agencies are typically run by someone appointed by the President and confirmed by the Senate, but the agency heads are typically supposed to be less political than the cabinet. For example, the President can fire the Secretary of State on a whim, and might do so to indicate a policy shift. On the other hand, the members of the Nuclear Regulatory Commission, for example, are appointed for fixed five year terms and shouldn't be removed for political reasons.
Regardless though, they are still part of the government.
The CIA reports to the DNI which is also selected by the Whitehouse and confirmed by the Senate. For independent, they sure have an awful lot of Executive and Legislative branch interaction for their top leadership.
Civilians or military, they're all Federal employees. Your analogies are simply wrong.
The funding is provided by the people VIA government. But the government doesn't participate in its administration nor does NASA participate in the execution of the Constitution.
There are A LOT of independent agencies that are owned or otherwise funded by the federal government but they do not participate in governance, and because of that, they are not PART of the federal government.
He made a small mistake, GET HIM! Everyone licks their lips
His examples of academia and DARPA were correct. The mentality of people here who can't wait to pile on and attack someone at the slightest error is sickening.
Calm down. The technology is from the 1990s. It's in the article:
" In the late 1990s, they began designing instruments for the next generation of satellites."
Also, enough with the empty comparisons between the US and "the world". There are many other countries that have developed "the most high-tech" devices too, you know? Like Plank from ESA, which is used to map the background radiation.
And regarding taxpayer dollars: do you know how long it takes for governmend-funded projects to complete? DECADES because the process is so slow precisely due to government red tape.
They never specified US government or US tax dollars. Merely that government and tax dollars funds science projects. You're reading into and attacking something that isn't there, then ended it on a smug and rude note.
>Calm down. The technology is from the 1990s. It's in the article:
The article states that instrument design began in the late 90's - not that that is what eventually made it into the satellites after two decades. They knew these satellites were approaching EOL and needed to be replaced so began R&D is how I read it.
>And regarding taxpayer dollars: do you know how long it takes for governmend-funded projects to complete? DECADES because the process is so slow precisely due to government red tape.
These are also often projects that would never be privately funded because there isn't necessarily a return on investment. They do, however, perform a public service. Maybe I'm naive - but I don't think weather satellites are a very good investment if one is looking for profit. Though I do admit I don't know how much one can make selling the data it produces.
Also, obviously all projects don't involve decades of work. The U.S. won WWII on two fronts in 4 years, put a person on the moon in less than ten, builds roads and bridges, rolled out the ACA ... and that's mostly just he big stuff.
Exactly, and moreover, it would be an incredible waste of resources not to share. Nice to see an example of the world scientific community being allowed to take the practical approach.
There is a benefit to countries and to the world when the people pool their money for something big.
Yes there is, at least sometimes. But, the choice to do that should be voluntary, otherwise it's an immoral and unjust situation. And given that the State is - at it's core - a mechanism rooted in violence, this kind of thing is not an appropriate role for government.
I wish they'd leave the 100mhz APT signals on these so kids around the country could build their own little "weather stations"... Being able to listen to those (I never had a wideband enough radio to see them) was one of the coolest things.
They've moved to Ghz band (Ku?) for weather several years back [1].
Great moves for science and I am happy that we are advancing tech, but an upgraded version of APT nowadays with how cheap tech is getting could have provided people in developing countries some better pictures of coming weather. Also could have kept the 10 year old inside of us with cheap 100mhz radios outside at night, listening to the satellites passing over our heads.
Perhaps my final statement was a bit glib, but personally I feel we are born into this world with a whole ton of obligations. We are nurtured, educated, protected. In many parts of the world, we are assigned inalienable rights just through the fact of being human beings. Whatever we do is founded on the hard work and achievements of others, but in most countries in the modern era very few claims are made on us in return.
In past ages we were expected, or even required to die or at least kill and risk death in defense of our society. Nowadays you can be brought up, educated to a decent standard, benefit from all of society's resources and benefits to the age of adulthood and then swan off to some other part of the world for the rest of our lives, obligation free. That's a pretty amazing deal, by historical or really any standards.
but personally I feel we are born into this world with a whole ton of obligations.
How in the world can one be born with obligations? Especially considering that you didn't even ask to be born to begin with? When you're born you certainly haven't had a chance to voluntarily enter into any contracts, or accept any burdens. What you're talking about sounds like slavery to me.
Laws and their enforcement are inherently coercive, but slavery is a specific thing distinguished by the fact that it is imposed through an asymmetric application of the law. Slaves and the owners are unreasonably assigned different rights under the law. That's the primary distinction that makes it immoral.
Do you feel unfairly and unreasonably constrained by the laws which oblige you to refrain from wantonly murdering and abusing those around you at will? Do you feel enslaved by those laws, or merely coerced?
Sorry to reply so late. But like you, it seems, I'm also interested in the tension between social welfare (education, medicine) and national borders. I'm saying this as someone who wants 10 million new immigrants to the USA per year.
The arguments shift, a few years back there was very widespread denial that the climate was changing. Following repeated record years now there is a reasonable consensus that the climate is changing (up to about 60% of conservatives in the US and UK), and people now seem to think this acceptance was always there. The next argument as you say is whether it is manmade, after that it will be about whether it is serious enough to need tackling, or about whether it's so serious that nothing can be done. Underlying this is the basic worry to avoid damaging quality of life (which is a perfectly reasonable concern) and the ideological motivation to avoid intervening in the economy. Beyond the motivated reasoning there is also the actual scientific argument about whether the climate sensitivity is 1C, or 2.5C, and whether therefore we should be taking moderate action or severe action. There's not much of a scientific debate about whether we should be doing nothing at all. Ironically in the Anglosphere there is a consensus for moderate action amongst the public, including the conservative public, but the political elites haven't yet caught up.
It's immoral and unjust that someone should grow up and live within the protective environment of a society, enjoying use of it's shared services and resources, while refusing to contribute towards the support of the services which make that environment possible.
Not when you were never asked if you wanted any of that stuff, and aren't given a choice as to whether or not to participate.
If you don't think it's a good deal, leave.
What, because a bunch of dead people drew some lines on a map and said "this is Whateveristan, and anybody inside these lines is bound to our decisions for all time"? Screw that. I'm with Thomas Paine who argued that there is no mechanism by which the dead can bind the living.
No, it isn't "fair". Voting is a sham. "Democracy" is just a euphemism for "mob rule" and "might makes right". As the old saying goes "Democracy is two wolves and a sheep voting on what to have for dinner. Liberty is a well armed sheep contesting the outcome."
The reality is, if you vote, but are on the losing side, you're just a minority suffering the Tyranny of the Majority[1]. The representative system we use in the US purports to try and avoid that, but it really doesn't.
and aren't given a choice as to whether or not to participate.
Interesting that you elided this bit in your response. Because this matters... no one is really given a choice as to whether or not to participate in "the system". If you work, the government presumes the authority to steal a portion of your income, no matter what decisions you make. And if you try too hard to not participate, men with guns will come and arrest or kill you.
And, no, before you say it, "leave" isn't a valid choice. See above about "Whateveristan". Given wherever one happened to be born, there's no reason one should have to leave in order to enjoy Liberty.
105 comments
[ 4.0 ms ] story [ 177 ms ] threadGhostbusters reference: http://vignette1.wikia.nocookie.net/villains/images/7/7f/Goz...
The R is just a sequence number for launching; S and T are slated next. If there is a failure before commissioning, it won't receive a numerical designation.
So the GOES-R name was just a happy accident I guess.
I think this is just a pattern of people seeing the complexity in their field, but not knowing anything about other fields and thus believing that they are doing the most awesome thing ever.
LHC guys will say their technology is the most complex one, Intel, the AI guys, and so on..., not to mention the global infrastructure required to run the Internet.
'Complexity', in this sense, doesn't mean 'it's rocket science' bus is a definable quantity.
They're referring to the total technological package, supercomputer forecasting, in addition to satellite building, lofting, and monitoring.
It's not a reference to the complexity of chaos theory. At least that's what I get from the context of the quote, which continues, “And we do it together as one species. It’s one of our greatest successes.” Made me get the warm fuzzies, sometimes we're not so shabby. :)
To be relevant, complexity should measure not the entire system but the number of variables at a specific point in the system. Rocket science isn't very complex as it is very compartmentalized. Different stages in the process involve discrete decisions divorced from others scheduled to occur at different times. So, weather forecasting wins because the number of variables at every decision is far greater. The flow-chart for a rocketry decision is more vertical whereas the chart for a weather forecast is more horizontal.
Where should I go now - not so much.
In terms of raw computing power I don't think there's any industry that utilizes as much super computing time as weather prediction. So there's also that.
For comparison, a dedicated AWS m4.4xlarge instance, which is equivalent to one of these nodes without the expensive Infiniband, is about $500/month. For an actual supercomputer node you can figure about double that.
So the total running cost (of just the hardware) for global weather prediction per month is probably around 30 million USD. Then you also have to pay people to manage the running of simulations, do the analysis, produce nice plots and animations for TV, etc.
Also, I Am Not A Weather Forecast Scientist, just working in a related field (fluid dynamics). So I could be wrong about stuff here.
As for shared models and data: I know Norway and the UK do a lot of this. I imagine so does the US and Canada/Mexico.
Also, keep in mind that weather modelling relies heavily on running an ensemble with random variations in initial/boundary conditions. So a lot of the computational cost comes from running hundreds of these simulations. My guess is that you don't run the large area 100 times, just a few times and then perturb it for all your local area simulations.
In our case, we have 1000-2000 instances (the number is elastic, because AWS) handling real-time ad auctions and related data processing. Since we are just one of the many players in the industry, and far smaller than the largest ones (GOOG, FB), I am pretty sure in aggregate more computing cycles are used to optimize ads than oil or weather.
Crazy.
Some statistics extracted from the article (written by a non-technical journalist, I suspect, so treat them accordingly): 5,200 servers, 115,000 CPUs, 3.8 petaflops, 27 petabytes.
Bear in mind that BP is only one oil company, and by no means the largest (you probably think Exxon is the largest, but Saudi Aramco is much bigger [2]).
[1] http://www.politico.com/sponsor-content/2015/12/journey-to-t...
[2] https://www.bloomberg.com/news/articles/2016-01-07/too-big-t...
Weather is both easy to measure and quite hard to predict and we've studied it for longer then most things. So we have a vast global array of sensors and ensembles of complicated models running constantly on supercomputers.
I hope we get to that level with biology and other fields at some point but we aren't there yet.
"Well, they said it would rain and it didn't..."
So, from me at least, thanks for all the hard work!
All I got from the actual information is a better resolution and faster analysis, nothing that would support the hyping.
Weather is incredibly complex and difficult to model and predict. We have multiple sources of truth to provide the input to our models: radar sites spaced hundreds of miles apart, balloons that are launched twice a day and hundreds of miles apart, METAR stations that record ground observations, satellite, etc. Satellite data is the closest you can get to seeing the entire atmosphere as an interconnected system with no gaps in data, but is limited by being a top-down view (better spectral information will help), somewhat low resolution, and 5-15 minute update intervals (storms can go from nothing to tornadic in 45 minutes).
CIMSS and other groups of researchers are doing great work with our current satellite data like identifying when supercells are likely to produce tornadoes and estimating instability in a column of the atmosphere. Increasing spectral information, resolution, and frequency will enable products to be refined and new products to come out - such as lightning detection, which currently relies on private networks.
Severe weather nowcasting is only one of the areas that will benefit from this, but the one that I'm familiar with as an amateur meteorologist. GOES-R has the potential to cut false alarm rate and increase warning times for severe weather. Weather nerds are rather excited, as this stands to be a powerful new tool.
GOES-R represents a nearly 2 orders of magnitude improvement over previous generations. Imagery will have higher spectral and spatial resolution, by significant margins, and provide coverage in new areas. Consider just one aspect: lightning observation, GOES-R will vastly increase the amount of data available. This is a huge amount of data that wasn't previously available and will no doubt be able to be used in improving weather modelling. However, lightning intensity has been recently discovered to be a key predictor in thunderstorm intensity and tornado formation so having that data available in real time will likely end up leading to better tornado predictions and many lives saved. And that's just one element, every aspect of the satellite is a huge leap over previous generations and will have a tremendous impact on weather forecasting and climate research for years to come.
Last i remember, all data transmitted by these satellites are not encrypted, anyone can build a receiver to gather the data to generate satellite images.
One day I hope to have time to do this myself.
http://www.rtl-sdr.com/rtl-sdr-tutorial-receiving-noaa-weath...
Found the beginning article to the series here:
http://www.rtl-sdr.com/setting-up-a-goes-weather-satellite-a...
Edit: For those who don't normally work with this stuff, I'm talking about not only correcting geospacial data - but also bringing the sensor data in line with a bunch of other sensors and past datasets. It quickly goes from "Oh, pretty pictures!" to "Do I really want to do this anymore? Can I ever trust data again?" That said, doing your own ground segment is cool - just know that there is a whole lot more work to do beyond actually receiving the radio transmissions.
https://earth.esa.int/web/sentinel/level-1-post-processing-a...
The U.S. military lacks critical funds at this point due to defunding, depending on how you define "critical". The service chiefs testified that they will not be able to defend the country unless they get more money.
http://breakingdefense.com/2016/09/four-service-chiefs-say-t...
Our technology advantage alone allows us to realistically probably coast for 6-8 years and still be matched only realistically by China. Russia has a few things that can match current US kit (their new tanks are excellent and they have some of the best anti-air kit in the world), but the Kremlin simply far and above lacks the cash to be in a position to afford to pay for a serious modernization. Obama's sanctions along with the massive drop in oil have put a dent in Russia's economy regardless of Putin's posturing. Sure they have very nice new jets, but they have precious few of them. We have nice new jets, better radars, and all of our older jets can spank the pants off of their older jets, etc, etc.
This is really just my anecdotal experience as a US Army veteran who was in OIF II 2003-2004.
I follow foreign policy issues closely, and I disagree.
> The current Order of Battle of the entire US Military calls for us to be able to fight two large scale wars simultaneously. We've done that arguably well/poorly with Iraq and Afghanistan.
I'm not sure what you mean by that usage of "Order of Battle". IIRC US military planning calls for them to fight one war, a medium-sized one IIRC, and keep a second medium-sized war in a holding pattern. Iraq and Afghanistan are small wars against very weak opposition whose best weapons are IEDs.
However, there isn't much here beyond claims with no basis. I'd be interested in seeing some evidence for it all.
The "Order of Battle" is US Military terminology to denote the composition of a military unit to respond to certain threats. It is most commonly used from Division / Brigade level and lower by the intelligence officers, but goes all the way up to the Pentagon & Department of Defense. An order of battle is designed to solve the question of, "How do I shape a military unit to best defeat the adversaries they're pledged or expected to fight against?". Absolutely nothing in the US's two war doctrine says for us to fight a medium war and be prepared for another medium sized war. Not sure what you think, but Iraq and Afghanistan are both big wars where we've simply de-emphasized "winning" them to optimize for less soldiers dying. The land mass is enormous in both and since a war is basically about owning a piece of land, it is easier to judge the size by the size of the land being attempted to liberate. Also, if we were fighting a larger conventional war (sans nukes) we would likely win/lose much quicker with much lower casualties. The asymetric warfare we're currently waging is something the US simply isn't very good at (nor is any conventional military really). What do you think?
For instance, I was a member of the US's first "Stryker Brigade Combat Team" (D Troop, 1/14th Cavalry, 2nd Brigade, 2nd Infantry Division out of Ft Lewis, Washington). Our Brigade of ~5500 soldiers replaced the 82rd Airborne division of ~19000 soldiers and did equally well due to our Order of Battle emphasizing speed (via the Strykers mainly) over sheer massive footprint and heavy slow armor. This was due to the enemy not being heavily armored and primarily using small pickup trucks "technicals". In this area, we pretty much dominated it as you'd have a fully loaded Stryker squad across the city (Tall'Afar or Mosul primarily) before anyone even realized it as they were much quieter than say a Hummvee.
Government doesn't do shit, except write the check. this all comes from acadamia, nasa, darpa, etc.
You're point that science is pushed forward by people pooling resources is absolutely correct, I just want to make it clear that Government has very little to do with it. This is just societ bolstering the resources of our greatest minds.
>nasa, darpa
I don't think you understand...
But they wouldn't. And you know it. Even the Science Philanthropy Alliance (http://www.sciencephilanthropyalliance.org/what-we-do/resour...), which funds researchers using privately-donated money, says that taxes are necessary to have scientific progress.
And yes, we are talking about semantics (mathematical sense). People couldn't give money to the DoD and accomplish the same thing, nor could they give money to DHS or any of the other Executive Departments.
NASA is an independent, federally funded, civilian agency. The part of it that is Federal, is its budget.
As opposed to the Department of Energy or Department of Education for example.
Indeed they are. Their budgets are paid for by taxpayers and established by laws passed in Congress and signed by the President. Their employees are federal employees, AFAIK.
> NASA for example doesn't actually rollup to any cabinet position.
The agencies report to the President, whatever the org chart is. The President chooses their administrators, and directs their operations (assuming she/he has time and interest).
DAPRA is in the Department of Defense.
Finally, I believe the CIA was an independent agency, like NASA, before the DNI was created about a decade ago. Was the CIA not actually part of the U.S. federal government?
https://www.hq.nasa.gov/office/hqlibrary/ic/faqs.html
DARPA is part of a department, the United States Department of Defense.
http://www.darpa.mil/
Both NASA and DARPA are as Federal as Federal gets and as govmint as govmint gets.
Full disclosure: my daddy worked for NASA on Mercury thru Apollo. He worked for the govmint.
The ONLY thing Federal about them is their funding.
There are "independent agencies" in the US Executive Branch, but these aren't independent of the entire US Government; they're just not part of a single Department. For example, NASA could have been part of the Department of Transportation (and thus report to the Secretary of Transportation), but it isn't and reports directly to the President.
Agencies are typically run by someone appointed by the President and confirmed by the Senate, but the agency heads are typically supposed to be less political than the cabinet. For example, the President can fire the Secretary of State on a whim, and might do so to indicate a policy shift. On the other hand, the members of the Nuclear Regulatory Commission, for example, are appointed for fixed five year terms and shouldn't be removed for political reasons.
Regardless though, they are still part of the government.
The CIA reports to the DNI which is also selected by the Whitehouse and confirmed by the Senate. For independent, they sure have an awful lot of Executive and Legislative branch interaction for their top leadership.
Civilians or military, they're all Federal employees. Your analogies are simply wrong.
Via government.
There are A LOT of independent agencies that are owned or otherwise funded by the federal government but they do not participate in governance, and because of that, they are not PART of the federal government.
The military does not participate in governance, by law. Is the U.S. military not part of the U.S. government? What about intelligence agencies?
It's an interesting definition, but I'm not sure that many people will accept it.
His examples of academia and DARPA were correct. The mentality of people here who can't wait to pile on and attack someone at the slightest error is sickening.
Calm down. The technology is from the 1990s. It's in the article:
" In the late 1990s, they began designing instruments for the next generation of satellites."
Also, enough with the empty comparisons between the US and "the world". There are many other countries that have developed "the most high-tech" devices too, you know? Like Plank from ESA, which is used to map the background radiation.
And regarding taxpayer dollars: do you know how long it takes for governmend-funded projects to complete? DECADES because the process is so slow precisely due to government red tape.
So, calm down boy.
>Calm down. The technology is from the 1990s. It's in the article:
The article states that instrument design began in the late 90's - not that that is what eventually made it into the satellites after two decades. They knew these satellites were approaching EOL and needed to be replaced so began R&D is how I read it.
>And regarding taxpayer dollars: do you know how long it takes for governmend-funded projects to complete? DECADES because the process is so slow precisely due to government red tape.
These are also often projects that would never be privately funded because there isn't necessarily a return on investment. They do, however, perform a public service. Maybe I'm naive - but I don't think weather satellites are a very good investment if one is looking for profit. Though I do admit I don't know how much one can make selling the data it produces.
Re-read the title sweetheart: it talks about the US, so it must obviously pertain to US gov't and US dollars.
From your profile:
>> Message To Self: It is okay to be wrong.
Yes, you are wrong, so calm down, kid.
Please stop. We don't treat other users this way on Hacker News, and we have to ban accounts that continue like this.
Yes there is, at least sometimes. But, the choice to do that should be voluntary, otherwise it's an immoral and unjust situation. And given that the State is - at it's core - a mechanism rooted in violence, this kind of thing is not an appropriate role for government.
Great moves for science and I am happy that we are advancing tech, but an upgraded version of APT nowadays with how cheap tech is getting could have provided people in developing countries some better pictures of coming weather. Also could have kept the 10 year old inside of us with cheap 100mhz radios outside at night, listening to the satellites passing over our heads.
1: https://en.wikipedia.org/wiki/Automatic_picture_transmission... & https://en.wikipedia.org/wiki/High-resolution_picture_transm...
I found them amazingly accurate. Now I know why. I am looking forward the future forecast provided by next gen satellite.
https://en.wikipedia.org/wiki/Somalia
In past ages we were expected, or even required to die or at least kill and risk death in defense of our society. Nowadays you can be brought up, educated to a decent standard, benefit from all of society's resources and benefits to the age of adulthood and then swan off to some other part of the world for the rest of our lives, obligation free. That's a pretty amazing deal, by historical or really any standards.
How in the world can one be born with obligations? Especially considering that you didn't even ask to be born to begin with? When you're born you certainly haven't had a chance to voluntarily enter into any contracts, or accept any burdens. What you're talking about sounds like slavery to me.
Do you feel unfairly and unreasonably constrained by the laws which oblige you to refrain from wantonly murdering and abusing those around you at will? Do you feel enslaved by those laws, or merely coerced?
Not if the law is no more than the collective extension to our individual, innate right to self-defense.[1]
[1]: http://bastiat.org/en/the_law.html
Not when you were never asked if you wanted any of that stuff, and aren't given a choice as to whether or not to participate.
If you don't think it's a good deal, leave.
What, because a bunch of dead people drew some lines on a map and said "this is Whateveristan, and anybody inside these lines is bound to our decisions for all time"? Screw that. I'm with Thomas Paine who argued that there is no mechanism by which the dead can bind the living.
You are indeed asked. You may be outvoted, but that's only fair. Laws are not immortal; they can be changed any time.
No, it isn't "fair". Voting is a sham. "Democracy" is just a euphemism for "mob rule" and "might makes right". As the old saying goes "Democracy is two wolves and a sheep voting on what to have for dinner. Liberty is a well armed sheep contesting the outcome."
The reality is, if you vote, but are on the losing side, you're just a minority suffering the Tyranny of the Majority[1]. The representative system we use in the US purports to try and avoid that, but it really doesn't.
[1]: https://en.wikipedia.org/wiki/Tyranny_of_the_majority
and aren't given a choice as to whether or not to participate.
Interesting that you elided this bit in your response. Because this matters... no one is really given a choice as to whether or not to participate in "the system". If you work, the government presumes the authority to steal a portion of your income, no matter what decisions you make. And if you try too hard to not participate, men with guns will come and arrest or kill you.
And, no, before you say it, "leave" isn't a valid choice. See above about "Whateveristan". Given wherever one happened to be born, there's no reason one should have to leave in order to enjoy Liberty.