Here near Boston at Alewife they have "Pedal-and-Park". It's a great big cage full of double-decker bike racks, with cameras on it, and a door that is locked and controlled by a (very easily obtained) access card. In the one instance of theft that I've heard of, they caught the thieves. See first and last photos: http://dr2chase.wordpress.com/2011/05/21/if-you-build-it-they-just-might-come/ And also, look at the induced demand -- for *bicycle parking* -- if you ever doubted that induced demand was possible.
Shift to another method of transportation. Motorcycle, carpool, bike, e-bike, transit, transit+folding bike, move, change jobs. The whole point is to attach the price to the thing that "costs" and let the market work it out.
There's not a direct connection between the price of the upgrade and the cost to use the road -- instead of funding through congestion charges on that particular road at rush hour, it's funded from taxpayer dollars. We do that because it's approximately a public good, but the accounting is really crappy (does it includes the external costs, from noise, crashes, pollution, oil spills, as well as impediment to travel perpendicular to the freeway?). If the goal is to reduce congestion, you can do that in a number of ways, and simply building more road is not necessarily the cheapest or most effective use of dollars. Standard microeconomic theory says, with much handwaving, that if you can tie the charge directly to the costs (i.e., to pay a fee to drive when demand is high and congestion is likely to result) then you will get the most efficient solution, as long as the transactions charges are not too high. Toll booths have transaction costs that are too high so we have traditionally not used congestion charging that much, but nowadays we can read license plates on the fly for cheap.
Bikes are an option. Either folding bikes that you take with you, or bikes at either end. Car-free friend of mine plays that game between Cambridge and Providence -- bike to Boston-Back-Bay on "good bike", take commuter rail to Providence, hop on "beater bike" at that end to get to work. "Beater bike" has survived over a year locked up in Providence. Full-sized bike on a train is more of a problem -- takes up space, and slows boarding.
But either way, use of a bike multiplies your tolerable-distance-from-station by a factor of 3 to 5.
That's an insanely stupid and costly idea. Even adding a normal-sized bicycle reduces the capacity of a train car by 2-3x (source: Caltrain bike cars, 40 bikes take the same space as 40 up-to-2-person seats). The effect of adding that bicycle to a boarding is a concern for getting the trains in and out of stations quickly enough.
Never mind that some of us have seen it in our own lifetimes? When you add freeway capacity, you create the (temporary) ability to live a little further from where you work, either to find a better job without moving, or to live in a somewhat cheaper house. In the case of a place like NYC or Boston where many people already take mass transit, making the freeway more attractive will pull people off of mass transit and onto the roads.
Perhaps "induced demand" would make a bit more sense if you looked that it the other way around. Do congestion and traffic jams make people more likely to drive, or less? I assume you would say "less" -- stuck in traffic sucks, and dinking around on surface streets (which jam up pretty quickly as soon as "everyone does it") is no fun, either. So why wouldn't adding capacity cause people to decide to drive more often?
Except that if we "accounted" honestly for congestion and accident costs, we would be far more willing to contemplate congestion charges. What if it turned out that the cheapest and most effective solution was simply to charge people to use the freeway, and use that money to subsidize bus transit to give people an alternative to the roads?
Note that there is no evidence in the fossil record of the Archangel Gabriel descending, and there is also no evidence of an ice age ever occurring with CO2 at current levels, never mind the levels we are projected to achieve later this century. It's much easier to assert the impossibility of those events.
There *IS* fossil evidence of rapid sea level rise, and we don't have a solid explanation for exactly how that happened -- meltwater, for example, requires an extraordinary amount of in-place melting (takes a tremendous amount of heat), and we have the additional fun of punching up CO2 at an unprecedented rate, so we cannot take past events as an upper bound on what we might observe.
There is the additional risk-unpleasantness of the arctic icecap melt completely outrunning theoretical predictions, and the obvious linkage between Greenland and the arctic ice cap. I'd be much more comfortable rejecting Hansen's claims as lacking a proper theoretical explanation if the proper theoretical explanation had not already been proven far too conservative in that region of the world. That's a huge problem.
And again, I'm not saying that Hansen is right -- I am saying that we don't have a comfortable level of certainty that he's wrong. I'm not normally a fan of exponential curve fits either. Unfortunately, he's a climate scientist, and I'm not (and as far as I know, you aren't either).
Perhaps you're comfortable with a higher level of risk, but I notice that you don't say that rapid rise was *certainly* due to a meltwater pulse, merely *likely*. Even if that event was certainly due to a meltwater pulse, that is not in itself proof that a similarly rapid glacial slide cannot occur now, given the unprecedented forcing. I do agree that it has to be a glacial slide -- melting in place would require a truly awesome amount of heat -- it takes far more heat to melt in place than is available merely from global warming (but there is plenty of heat for melting ice once it is in the ocean).
An exponential fit is aggressive, I agree, but do you have a proof of impossibility that does not rely on bluster?
We don't know this with the certainty that we'd like, certainly not in the case of Greenland. Antarctic, more likely that will take centuries, I'm not sure it's supposed to even melt 100% in the likely futures. In the case of Greenland, there's fossil/isotopic evidence of very rapid sea level rise in the past (multiple meters per century), and the current measurements of melt rate show that it is accelerating. http://boingboing.net/2011/07/21/why-we-need-to-keep.html
Exponential may have the best fit, but exponential is a very aggressive estimate. Quadratic is a conservative simple choice, and it's not completely divorced from theory -- the short-term "obvious" additional heat input is linear in reduced albedo, which gives you quadratic.
I'm not terribly happy at just saying "screw the theory, look at the curve fitting", but theory (such as we know it, which is a lot of the problem here) has severely underpredicted the melting.
One approach looks only at ice volume measurements, and explicitly ignores theory because the existing theoretical models failed to predict anything like the ice loss that we observed. Using the simplest accelerating-curve-fit, we get first ice free in September 2017, and six months per year ice free by 2025.
Have a look at some of the stuff at ledsupply.com. Their stuff tends to be good. Newer CREE (XPG, MKR if they've got it) is what you want, you need to take some care in heat sinking. They sell drivers, also, both DC/DC and AC/DC, some of them dimmable from an external signal, some of them (I think) dimmable from a wall dimmer. Here's an example of how you work with their stuff: http://dr2chase.wordpress.com/2011/02/20/undercabinet-lights-basement-kitchen/
Compared to the light from an incandescent bulb, LEDs have plenty low heat output, and I implied (but didn't say it plainly enough) that you might distribute the LEDs in smaller units than is current practice with incandescent fixtures.
3W LEDs (one amp into an XRE or XPG) need something like 9 square inches of aluminum stock in free air, 1W gets by with plenty less. 15 watt needs more, obviously. But for a standard ceiling, if you were putting discrete 3W LEDs up there, you could cut a hole in the drywall, use about 6-9 inches of 2" aluminum flat stock, and just epoxy the LED on to that. Put 1/2 standoffs on the bottom side of the flat stock, slip the whole thing in to the hole (LED side down, obviously), and you've taken care of cooling. Find some photogenic way to put a diffuser over the hole you cut in the ceiling, and you're done.
I think that LEDs might be doing better than 40% right now, too. I've read, somewhere, that best we can do with "white" is about 300 lumens per watt, and CREE is apparently selling (I saw your note below) 200 lumens/watt. The theoretical max lumens for a single wavelength is more than that, but that's green, not white.
Rewiring and adding or changing fixtures typically costs a little bit less than demo and rebuild of the whole house. We've had trained professionals do it for us (new stove), and I've added circuits all by myself from basement panel all the way up to a 2nd-floor bathroom. Lighting under cabinets with LEDs is dead simple, and the heat output from individual LEDs is so low and the fixtures so light-weight that you can make do with a very low-tech "fixture" (e.g., a few inches of aluminum flat stock).
Here, some pictures worth quite a few words: http://dr2chase.wordpress.com/2011/02/26/led-color-rendering/ (From a similar discussion, here on Slashdot, about a year ago). It's my opinion that the camera magnifies the differences, which means that the mixed LED case (cool+neutral+warm) looks really good to your eyes in practice. I think it would be good to fit some diffusers over the point sources, because sometimes you do get an "interesting" multi-shadow effect.
I use a bunch of good-quality (*) neutral CREE LEDs under our kitchen cabinets, and that looks fine, and my better half does not complain (and she does note that most fluorescent bulbs look like shit). Here: http://dr2chase.wordpress.com/2008/10/19/more-undercabinet-lights/
(*) for a given manufacturer, so a "CREE neutral white" might be one of 2, 3, or even 6 different bins (I just checked across XML, XPG, XRE products).
It has an enormous lot to do with the quality of the power supply components and how hot the bulb gets (this is also true of LEDs). Comment on TFA mentions this -- electrolytic capacitors have a lifetime that is very sensitive to heat, and can be quite short.
The main flaw with these energy star standards is that they too heavily weight towards backwards compatibility -- if, say, someone came up with a new way of packaging LEDs into new construction, where the lights and the power supplies were decoupled (one power supply, many little lights), the energy star standard would be simply unable to evaluate it -- it's not a "100W replacement", it doesn't fit into a standard fixture, etc. And there's good technical reasons to do it that way -- spreading out the lights simplifies the cooling, getting the power supply away from the lights helps with keeping those components cool, etc.
What makes you so sure that it is all about the revenue? There's some evidence that revenue is a factor ( http://query.nytimes.com/gst/fullpage.html?res=F70713FC3E580C718CDDA00894DF404482 -- "Political Economy at Any Speed: What Determines Traffic Citations?"), but it's not a huge one. Being from out-of-town give you a 10% higher risk of a ticket versus a warning, and driving through a town that was tight for money gives you a 28% higher chance of a ticket versus a warning. 28% sounds like a lot, but there's that other 72% to think about.
If it's a residential area, you should not go even 1mph over the speed limit. Crash risks to pedestrians grow dramatically in the 20-40mph range.
And why shouldn't drivers follow traffic laws to the last detail? It's not as if cars are harmless; tens of thousands of drivers and passengers are killed each year, thousands of pedestrians are killed each year, and the injury rate is an order of magnitude higher.
Why would I believe you, over, say, Krugman and Delong? Both those guys back in 2009 rather loudly proclaimed that we needed economic stimulus, that Obama's plan was not large enough, and that hyperinflation would not result. They were right on the first point -- compared to England, we're doing well; right on the second point -- the economy still sucks, we needed a bigger boost; and right on the third point -- we see practically no inflation. Can you point to any other economist with a better track record at predicting economic behavior in this recession?
History, for example our premature attempt to cut the deficit back in 1937 (which caused a slump in the recover from the Great Depression) also suggests that deficits are not that bad. We ran a monster deficit to fight WW2, and that finally put an end to the Great Depression.
And furthermore, the market, allegedly the decider, says that our deficit spending is NOT a bad idea -- if it were, nobody would loan money to our government at the ridiculously low rates that are currently in force.
Why would you think this? Look at the interest rates on government debt. Apparently there's no shortage of "retarded" people, some of them with quite a lot of money. If someone's willing to loan us money at lower interest rates than expected inflation, I think we should borrow.
And why is your reply so insulting and content-free?
Slashdot Mirror
Here near Boston at Alewife they have "Pedal-and-Park". It's a great big cage full of double-decker bike racks, with cameras on it, and a door that is locked and controlled by a (very easily obtained) access card. In the one instance of theft that I've heard of, they caught the thieves. See first and last photos: http://dr2chase.wordpress.com/2011/05/21/if-you-build-it-they-just-might-come/
And also, look at the induced demand -- for *bicycle parking* -- if you ever doubted that induced demand was possible.
Shift to another method of transportation. Motorcycle, carpool, bike, e-bike, transit, transit+folding bike, move, change jobs. The whole point is to attach the price to the thing that "costs" and let the market work it out.
There's not a direct connection between the price of the upgrade and the cost to use the road -- instead of funding through congestion charges on that particular road at rush hour, it's funded from taxpayer dollars. We do that because it's approximately a public good, but the accounting is really crappy (does it includes the external costs, from noise, crashes, pollution, oil spills, as well as impediment to travel perpendicular to the freeway?). If the goal is to reduce congestion, you can do that in a number of ways, and simply building more road is not necessarily the cheapest or most effective use of dollars. Standard microeconomic theory says, with much handwaving, that if you can tie the charge directly to the costs (i.e., to pay a fee to drive when demand is high and congestion is likely to result) then you will get the most efficient solution, as long as the transactions charges are not too high. Toll booths have transaction costs that are too high so we have traditionally not used congestion charging that much, but nowadays we can read license plates on the fly for cheap.
Bikes are an option. Either folding bikes that you take with you, or bikes at either end. Car-free friend of mine plays that game between Cambridge and Providence -- bike to Boston-Back-Bay on "good bike", take commuter rail to Providence, hop on "beater bike" at that end to get to work. "Beater bike" has survived over a year locked up in Providence. Full-sized bike on a train is more of a problem -- takes up space, and slows boarding.
But either way, use of a bike multiplies your tolerable-distance-from-station by a factor of 3 to 5.
That's an insanely stupid and costly idea. Even adding a normal-sized bicycle reduces the capacity of a train car by 2-3x (source: Caltrain bike cars, 40 bikes take the same space as 40 up-to-2-person seats). The effect of adding that bicycle to a boarding is a concern for getting the trains in and out of stations quickly enough.
Never mind that some of us have seen it in our own lifetimes? When you add freeway capacity, you create the (temporary) ability to live a little further from where you work, either to find a better job without moving, or to live in a somewhat cheaper house. In the case of a place like NYC or Boston where many people already take mass transit, making the freeway more attractive will pull people off of mass transit and onto the roads.
Similarly, when the roads get screwed up you see a reduction in auto traffic and a surge in mass transit use (for example, by an earthquake, as happened in the SF area after the Loma Prieta quake -- http://en.wikipedia.org/wiki/1989_Loma_Prieta_earthquake#Effects_on_transportation ) .
Perhaps "induced demand" would make a bit more sense if you looked that it the other way around. Do congestion and traffic jams make people more likely to drive, or less? I assume you would say "less" -- stuck in traffic sucks, and dinking around on surface streets (which jam up pretty quickly as soon as "everyone does it") is no fun, either. So why wouldn't adding capacity cause people to decide to drive more often?
Except that if we "accounted" honestly for congestion and accident costs, we would be far more willing to contemplate congestion charges. What if it turned out that the cheapest and most effective solution was simply to charge people to use the freeway, and use that money to subsidize bus transit to give people an alternative to the roads?
Note that there is no evidence in the fossil record of the Archangel Gabriel descending, and there is also no evidence of an ice age ever occurring with CO2 at current levels, never mind the levels we are projected to achieve later this century. It's much easier to assert the impossibility of those events.
There *IS* fossil evidence of rapid sea level rise, and we don't have a solid explanation for exactly how that happened -- meltwater, for example, requires an extraordinary amount of in-place melting (takes a tremendous amount of heat), and we have the additional fun of punching up CO2 at an unprecedented rate, so we cannot take past events as an upper bound on what we might observe.
There is the additional risk-unpleasantness of the arctic icecap melt completely outrunning theoretical predictions, and the obvious linkage between Greenland and the arctic ice cap. I'd be much more comfortable rejecting Hansen's claims as lacking a proper theoretical explanation if the proper theoretical explanation had not already been proven far too conservative in that region of the world. That's a huge problem.
And again, I'm not saying that Hansen is right -- I am saying that we don't have a comfortable level of certainty that he's wrong. I'm not normally a fan of exponential curve fits either. Unfortunately, he's a climate scientist, and I'm not (and as far as I know, you aren't either).
Not sure this is right for you, but it was once upon a time designed to match some of your buzzwords.
http://sourceforge.net/apps/trac/reddwarf/
Perhaps you're comfortable with a higher level of risk, but I notice that you don't say that rapid rise was *certainly* due to a meltwater pulse, merely *likely*. Even if that event was certainly due to a meltwater pulse, that is not in itself proof that a similarly rapid glacial slide cannot occur now, given the unprecedented forcing. I do agree that it has to be a glacial slide -- melting in place would require a truly awesome amount of heat -- it takes far more heat to melt in place than is available merely from global warming (but there is plenty of heat for melting ice once it is in the ocean).
An exponential fit is aggressive, I agree, but do you have a proof of impossibility that does not rely on bluster?
We don't know this with the certainty that we'd like, certainly not in the case of Greenland. Antarctic, more likely that will take centuries, I'm not sure it's supposed to even melt 100% in the likely futures. In the case of Greenland, there's fossil/isotopic evidence of very rapid sea level rise in the past (multiple meters per century), and the current measurements of melt rate show that it is accelerating. http://boingboing.net/2011/07/21/why-we-need-to-keep.html
Exponential may have the best fit, but exponential is a very aggressive estimate. Quadratic is a conservative simple choice, and it's not completely divorced from theory -- the short-term "obvious" additional heat input is linear in reduced albedo, which gives you quadratic.
I'm not terribly happy at just saying "screw the theory, look at the curve fitting", but theory (such as we know it, which is a lot of the problem here) has severely underpredicted the melting.
One approach looks only at ice volume measurements, and explicitly ignores theory because the existing theoretical models failed to predict anything like the ice loss that we observed. Using the simplest accelerating-curve-fit, we get first ice free in September 2017, and six months per year ice free by 2025.
http://earlywarn.blogspot.com/2012/08/more-on-arctic-sea-ice-volume.html
Have a look at some of the stuff at ledsupply.com. Their stuff tends to be good. Newer CREE (XPG, MKR if they've got it) is what you want, you need to take some care in heat sinking. They sell drivers, also, both DC/DC and AC/DC, some of them dimmable from an external signal, some of them (I think) dimmable from a wall dimmer. Here's an example of how you work with their stuff: http://dr2chase.wordpress.com/2011/02/20/undercabinet-lights-basement-kitchen/
Compared to the light from an incandescent bulb, LEDs have plenty low heat output, and I implied (but didn't say it plainly enough) that you might distribute the LEDs in smaller units than is current practice with incandescent fixtures.
3W LEDs (one amp into an XRE or XPG) need something like 9 square inches of aluminum stock in free air, 1W gets by with plenty less. 15 watt needs more, obviously. But for a standard ceiling, if you were putting discrete 3W LEDs up there, you could cut a hole in the drywall, use about 6-9 inches of 2" aluminum flat stock, and just epoxy the LED on to that. Put 1/2 standoffs on the bottom side of the flat stock, slip the whole thing in to the hole (LED side down, obviously), and you've taken care of cooling. Find some photogenic way to put a diffuser over the hole you cut in the ceiling, and you're done.
I think that LEDs might be doing better than 40% right now, too. I've read, somewhere, that best we can do with "white" is about 300 lumens per watt, and CREE is apparently selling (I saw your note below) 200 lumens/watt. The theoretical max lumens for a single wavelength is more than that, but that's green, not white.
Damn. I didn't know that was shipping yet as a product. And I see that they do the mixed-spectrum right on the die, and get very nice results.
Yes. Yes.
http://www.ecaa.ntu.edu.tw/weifang/led/Effects%20of%20Frequency%20and%20Duty%20Ratio%20on%20the%20Growth%20of%20Potato%20Plantlets%20In%20Vitro%20Using%20Light-emitting%20Diodes%20--%20Jao%20and%20Fang%2039%20(2)%20375%20--%20HortScience.htm
Rewiring and adding or changing fixtures typically costs a little bit less than demo and rebuild of the whole house. We've had trained professionals do it for us (new stove), and I've added circuits all by myself from basement panel all the way up to a 2nd-floor bathroom. Lighting under cabinets with LEDs is dead simple, and the heat output from individual LEDs is so low and the fixtures so light-weight that you can make do with a very low-tech "fixture" (e.g., a few inches of aluminum flat stock).
Here, some pictures worth quite a few words: http://dr2chase.wordpress.com/2011/02/26/led-color-rendering/
(From a similar discussion, here on Slashdot, about a year ago).
It's my opinion that the camera magnifies the differences, which means that the mixed LED case (cool+neutral+warm) looks really good to your eyes in practice. I think it would be good to fit some diffusers over the point sources, because sometimes you do get an "interesting" multi-shadow effect.
Here's what those same LEDs looked like in construction: http://dr2chase.wordpress.com/2011/02/20/undercabinet-lights-basement-kitchen/
I use a bunch of good-quality (*) neutral CREE LEDs under our kitchen cabinets, and that looks fine, and my better half does not complain (and she does note that most fluorescent bulbs look like shit). Here: http://dr2chase.wordpress.com/2008/10/19/more-undercabinet-lights/
(*) for a given manufacturer, so a "CREE neutral white" might be one of 2, 3, or even 6 different bins (I just checked across XML, XPG, XRE products).
It has an enormous lot to do with the quality of the power supply components and how hot the bulb gets (this is also true of LEDs). Comment on TFA mentions this -- electrolytic capacitors have a lifetime that is very sensitive to heat, and can be quite short.
The main flaw with these energy star standards is that they too heavily weight towards backwards compatibility -- if, say, someone came up with a new way of packaging LEDs into new construction, where the lights and the power supplies were decoupled (one power supply, many little lights), the energy star standard would be simply unable to evaluate it -- it's not a "100W replacement", it doesn't fit into a standard fixture, etc. And there's good technical reasons to do it that way -- spreading out the lights simplifies the cooling, getting the power supply away from the lights helps with keeping those components cool, etc.
What makes you so sure that it is all about the revenue? There's some evidence that revenue is a factor ( http://query.nytimes.com/gst/fullpage.html?res=F70713FC3E580C718CDDA00894DF404482 -- "Political Economy at Any Speed: What Determines Traffic Citations?"), but it's not a huge one. Being from out-of-town give you a 10% higher risk of a ticket versus a warning, and driving through a town that was tight for money gives you a 28% higher chance of a ticket versus a warning. 28% sounds like a lot, but there's that other 72% to think about.
If it's a residential area, you should not go even 1mph over the speed limit. Crash risks to pedestrians grow dramatically in the 20-40mph range.
And why shouldn't drivers follow traffic laws to the last detail? It's not as if cars are harmless; tens of thousands of drivers and passengers are killed each year, thousands of pedestrians are killed each year, and the injury rate is an order of magnitude higher.
Why would I believe you, over, say, Krugman and Delong? Both those guys back in 2009 rather loudly proclaimed that we needed economic stimulus, that Obama's plan was not large enough, and that hyperinflation would not result. They were right on the first point -- compared to England, we're doing well; right on the second point -- the economy still sucks, we needed a bigger boost; and right on the third point -- we see practically no inflation. Can you point to any other economist with a better track record at predicting economic behavior in this recession?
History, for example our premature attempt to cut the deficit back in 1937 (which caused a slump in the recover from the Great Depression) also suggests that deficits are not that bad. We ran a monster deficit to fight WW2, and that finally put an end to the Great Depression.
And furthermore, the market, allegedly the decider, says that our deficit spending is NOT a bad idea -- if it were, nobody would loan money to our government at the ridiculously low rates that are currently in force.
Why would you think this? Look at the interest rates on government debt. Apparently there's no shortage of "retarded" people, some of them with quite a lot of money. If someone's willing to loan us money at lower interest rates than expected inflation, I think we should borrow.
And why is your reply so insulting and content-free?
That's why you ride those employer-subsidized Google bicycles. Uh-oh.....