Slashdot Mirror
Weather Monitoring Frequencies Subject to Pollution
jd writes "In a case of technology vs. technology, the ICU (the body governing the use of radio frequencies around the globe) has been asked to secure radio frequences used for weather monitoring. In-car radar, mobile phones and other commercial and military applications are now using these same frequencies. However, weather satellites can't simply be re-tuned. There is only one very narrow band that detects water vapor but not liquid water, for example. This frequency has been sold to developers of car radar systems. The more this happens, the less useful weather radar and weather satellites will be. The noise will simply swamp the data, making what is collected useless. The article doesn't give a 'doomsday' timeframe, when we'll have no better ability to forecast the weather than they did in the 1800s, but that is what they are talking about."
Question to those who might know this - assuming the ITU agrees to these restrictions, how would they enforce them? The radar frequency was presumably sold by a national agency (a la FCC) which is clearly making money off the sale and doesn't seem to care about the reasons. So how would the ITU go about forcing them to behave?
Well when I feel like cheating and not drawing up my own weather forecasts. I cheat by going and type "San Francisco Forecast Discussion" into Google or Alta Vista and read that. It is pretty accurate and they even translate the weather-ese for you. If the term appears in browser link colour just click on it and it will be translated in a nice little box for you. It isn't that hard to learn to do basic forecasting. I have managed to teach it to seventh to ninth graders, who are always cheered when they beat the local weather guy. But unless Nexsat and NPOESS prove to be a real boon we really do need those freqs and we should get 'em back. I could go on a lot longer about this...But I will see if anyone is interested first.
HFSSK
I've seen somewhere that if you say "Tomorrow will have the same weather as today", you will have a probability of 73% to be correct. The meteorologists calculate the weather by creating a stochastic model, then simulate it maaany times in order and then get a kind of average weather from all the simulations. The probability of them being correct is about 78%...
I believe the reason is that you don't want to get reflections from objects far away. A good way of doing this is to use frequencies that dampens the signal alot.
A kid at my son's school collected and analyzied common RSS weather feeds [tech-recipes.com] for a science project.
/.
He collected the data and used it to judge how accurate the weatherman's predictions were.
Weather reports != Meteorologist predictions.
You would be amazed at how often and by how far the reports differ from what the meteorologists have predicted.
A bunch of random RSS feeds are going to tend to be inaccurate. Your kid confirmed that - kudos to him - sounds like a great project.
But hardly worth mentioning on
My pics.
I do a lot of forecasting based on computer models. You current accurately predict specifics, but you can work out trends quite accurately. Like I can look at time 3 days out and see that there is potential for severe thunderstorms. That doesn't say that they'll happen, but there's a potential for them. It's never absolute, but you can get an idea. With a combination of computer models and local knowledge you can get pretty good, but don't expect to be able to say exactly what the temperature will be, or if a particular locality will get rain. I think half the issue is that a lot of weather presenters don't really understand what they're talking about and predict it as if it's absolute truth.
legitimately in the first place.
The spectrum is a natural resource which belongs to everyone equally. No one, and no government, has any right to "own," or sell what belongs to everyone. The only legitimate role of government in spectrum is to regulate it to maximize the public good. It is false to assume that "selling" spectrum provides any public benefit at all (although those who support that would argue that increasing government revenue is somehow "good.")
Spectrum allocation is a large, time-varying, multivariable optimization problem. This document is an outline of some of the service requests/requirements, and how they need to mesh with each other, present and future technology availability, and physical limitations (like attenuation due to water at 24 GHz). Note that this document is only U.S. interests; every other country has a similar list, and all have to be coordinated. It's like the guy who goes into a store with three lists: What he wants to buy, what he needs to buy, and what he can afford to buy. Compromise is the name of the game, and reasonable people will make reasonable tradeoffs differently.
The radar this article is discussing is a proposed future use of 24 GHz for collision-avoidance radar in passenger cars. 24 GHz is a popular frequency choice for short-range applications like this specifically because of the atmospheric attenuation. Note that the attenuation at 24 GHz, while higher than at other nearby frequencies, is still relatively low, only a few tenths of a dB per kilometer (although much higher in rain). This makes 24 GHz a good compromise for short-range devices on the Earth's surface, especially low-powered devices with very directional antennas pointed horizontally, away from satellites. (A better choice from this standpoint would be the oxygen absorption band at 60 GHz, and there is indeed another radar band there.)
Meterologists are merely expressing their concern over how their measurements will be corrupted if millions of car radars are in operation, and their cumulative power is enough to be detected by their sensors. My personal opinion, however, is that 24 GHz is too low of a frequency to make a market-successful car radar; the antennas are too big. I think 60 or 77 GHz is a better bet; if so, that would preserve 24 GHz for water vapor measurements.
In general, though, the interests of meterologists and others performing microwave sensing of the earth should be considered in the frequency allocation process; the publicity due to this article is one way of accomplishing this.
Uh, no.
Chaos theory allows a scientist to identify which systems are chaotic, and which are not.
Under certain conditions, it may be possible to coerce a system back into predictability. Obviously, the practical application of such coercion is limited when one is studying global weather systems. On the other hand, if a heart starts beating in a chaotic fashion, and if that heart is equipped with a pacemaker, the heartbeat can be corrected with a mild electric shock. The timing and voltage of that shock can optimally determined with chaos theory. Similarly, some industrial processes might be amenable to such corrective measures.
People tend to view weather accuracy on a micro scale, but meteorologists can only work on a macro scale. If you look at satellite images from Canada's weather service, you can see that the city you live in is less than 1/20th the size of a pin-prick relative to a weather system.
When a weather report going out 24 hours into the future says it will rain, it WILL rain... just not perhaps overtop of your little pin-prick. Considering the complexity of weather, realistically how effective can we expect a 3-day outlook to be?
All I'm saying is think big, and you'll realize that meteorology is an inexact science, but is so very valuable as a critical service. Nobody will ever be able to 'stick their head out the window' and have a hope of predicting anything. Meteorologic science deserves a hell of alot more credit than the farmer's almanac.
A few questions for you, ApharmdB. I'm just curious.
Your radiometer was ground based correct? Were you looking directly at down-welling radiation where a drop in your signal output corresponded with an increase in water vapor density?
In order to use one antenna, it sounds like you had two tuned RF Front ends, one for each frequency. Did you use a standard heterodyne receiver architecture? What was your base-band bandwidth prior to your power detector or did you digitize? Polarization?
Sorry for all the questions. I've built a few radiometer systems before and I've always been curious about the different architectures and relative performance of them.
Related to the article, it is vital that we maintain an interference free spectrum around important atmospheric and astronomical bands. This could very easily become of greater importance than even BPL interference.