8 ms·

“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 com

by 21 10y ago

“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.

matt4077 10y ago
They're speaking of Chaos Theory, which was invented/discovered for weather prediction. 'Complexity', in this sense, doesn't mean 'it's rocket science' bus is a definable quantity.
- igravious 10y ago
  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. :)
- sandworm101 10y ago
  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.
  - kilotaras 10y ago
    Where should I go with my dog in 7 days can be argued that more complex. Where should I go now - not so much.
dmix 10y ago
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.
- tonfa 10y ago
  Out of curiosity, I wonder how it compares with computer power used for ad placement prediction.
  - semi-extrinsic 10y ago
    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.
    - detaro 10y ago
      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?
      - semi-extrinsic 10y ago
InclinedPlane 10y ago
All this proves is that you have absolutely no understanding of the actual work that goes into weather forecasting.
madengr 10y ago
I'm an RF EE; I have complex signals going into complex impedance. It don't get more complex than that.
- SEJeff 10y ago
  Curious, is that harder than fluid dynamics?
  - madengr 10y ago
    Ha ha, I intended it as a joke. Probably not. I'd say Navier Stokes equations are more difficult to apply than Maxwells equations.
SEJeff 10y ago
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.
micro_cam 10y ago
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.
- ethbro 10y ago
  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!
  - adolph 10y ago
    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.

by 21 10y ago

matt4077 10y ago

They're speaking of Chaos Theory, which was invented/discovered for weather prediction. 'Complexity', in this sense, doesn't mean 'it's rocket science' bus is a definable quantity.

igravious 10y ago
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. :)
sandworm101 10y ago
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.
- kilotaras 10y ago
  Where should I go with my dog in 7 days can be argued that more complex. Where should I go now - not so much.

dmix 10y ago

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.

tonfa 10y ago
Out of curiosity, I wonder how it compares with computer power used for ad placement prediction.
- semi-extrinsic 10y ago
  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.
  - detaro 10y ago
    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?
    - semi-extrinsic 10y ago

InclinedPlane 10y ago

All this proves is that you have absolutely no understanding of the actual work that goes into weather forecasting.

madengr 10y ago

I'm an RF EE; I have complex signals going into complex impedance. It don't get more complex than that.

SEJeff 10y ago
Curious, is that harder than fluid dynamics?
- madengr 10y ago
  Ha ha, I intended it as a joke. Probably not. I'd say Navier Stokes equations are more difficult to apply than Maxwells equations.

SEJeff 10y ago

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.

micro_cam 10y ago

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.

ethbro 10y ago
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!
- adolph 10y ago
  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.