If we had the wherewithal to move to a different (habitable) planet, I suspect that we could also deflect asteroids, break up tsunamis, etc.
We have the capability of colonising other areas of the Solar System right now. A mission to the right sort of nearby resource-rich asteroid could 'bootstrap' our exploration and colonisation of the rest of the Solar System and beyond.
Defecting large asteroids may be possible in the near future, if we have enough warning, but the idea that we could break up tsunamis or prevent thousands of cubic kilometres of lava erupting from a supervolcano is absurd - it would require energies and technologies beyond anything we can currently conceive of.
Here's a clue to the solar system: There's nothing there within our reach that'll save us when we finally fuck up the planet irreparably. Get over the concept that we could move and start trying to fix the Earth first.
Actually, we could move pretty easily if we really wanted to; we are planning to return to the Moon in the near future, and hopefully some sort of base will be established at some point.
The problem is the the Earth isn't fixable. Even if we don't cause long-term (from our point of view) global warming, then eventually we are going to be hit by an asteroid, or there will be a supervolcano somewhere, or a mega-tsunami event. These would have effects way beyond anything we could achieve (the dinosaur-killer asteroid hit with the force of a million nuclear weapons!).
The idea that we could survive as we are long-term on this planet is totally unreasonable.
It's doing the same run-time cached-interpretation thing as python or anything else.
No it isn't.
Here is the full details of what happens when a Java program starts on a modern JVM.
Initially, a byte code interpreter starts running the program. This is reasonably fast (like Python), but nowhere near the performance of optimised native code. At the same time, a background thread is started which profiles the byte code and finds areas which are taking up the most time. These areas of the code are analysed in a lot of detail, and then aggressively optimised, with variable re-arrangements, method inlining, loop unrolling and (where possible) bounds checking removal. Then - and this is the key stage - the byte code is translated to highly optimised native code. When I say 'highly optimised' that is exactly what I mean. Amongst other things, the JVM can analyse processor pipelines and re-arrange the order of machine code instructions for optimal performance, and for most math operations, the JVM will embed processor instructions inline.
So, Java is not doing interpretation, at least not on any code that takes up time. It is running native code.
However, this takes time, which is why many benchmarks that run for just a few seconds don't show good performance for Java.
If you run a Java with the '-server' switch - which means do a lot of optimisation at the initial expense of response - then you will find that within a few seconds your program could well give close or equal performance to the output of many C compilers.
I think the difference from Python should now be clear.
Why does the performance of python need to be any worse than that of Java?
Because Java byte code is translated into high-performance native code, unlike Python.
A python script will have bytecode generated the first time and then run as high performance native code on any platform.
No interpreted byte code runs at a performance equivalent to native code on any platform. So say this is the case is simply wishful thinking. This is why a huge amount of effort is put into JIT compiling of these byte codes to native code. There are projects under way to get this working for Python.
I'm pretty sure a lot of the processing was direct python. And some languages make it easier to combine than others.
I'm sure it was, but if you read back, I was comparing Java with pure Python code. The point was that to get good performance the core parts had to be C. This dropping through to another language is inescapably messy.
There's no reason you couldn't do it in pure python - there's just no point.
Of course there is a reason you couldn't do this practically in pure Python - performance.
So they've written two copies of it because Java people refuse to reuse existing libraries?
No, they have written two copies of it because having to configure and re-build C source code on every platform on which you want to deploy is tedious and outdated. The Java implementation can be shipped as byte code and will then run as good performance native code on any platform.
My point is that python is just as suitable for image processing as Java, and usually more productive.
But Python isn't used for image processing! In the examples you give, Python is being used as a layer above C that does image processing. Python is a great language to wrap around C code for things like this, but to claim that Python is doing the processing is to mislead.
This is in stark contrast to Java. The standard Java API for image processing is JAI (Java Advanced Imaging). Although it can use C libraries if they are available it can (if these libraries or not present) also use a Java-only implementation which gives good performance.
No matter what the advantages of Python (and there are many) to claim that it is equivalent to Java in this area is simply incorrect.
If you want high-performance numerical code Java isn't that suitable either, you want C or fortran at a minimum and probably hand-coded assembler.
Have you researched Java's numerical performance recently? There is no need for C or Fortran. Java produces high-performance assembler that does a good job with numerical work.
But I've seen plenty of serious image processing done in python, remember the recent jigsaw solving story?
That is not serious image processing done in python. That is python using C. I used the phrase raw python. Any language can 'spoof' speed by calling out to a faster one.
No need to be sorry, I'm glad to hear it. I'm a Java fan too. I'm stuck on Java 1.4 at the moment though, since we use it at work, so I haven't had time to test startup/Swing performance of the new versions. As soon as I'm finished with my math studies I'll install the recent versions at home and start developing.:-)
I don't know what they have done in Java 1.5 to speed things, but it really works - I am finding JEdit starting faster than typical equivalent KDE applications - like Kate (3 or 4 seconds).
Also, the look and feel of Swing has been changed - it was pretty awful in previous versions, but now has a sort of 3D shaded light blue look.
Best performing - SWT probably, but it is less portable.
SWT has recently had reports of performance problems on some platforms - Linux in particular. As Swing now uses hardware acceleration by default where possible (OpenGL,DirectX) the difference in performance between SWT and Swing is usually negligible. The fact that Swing is built-in to all JRE distributions gives it huge advantage.
Both still have the problem of JVM startup time though (another problem Sun is looking at, they are currently testing a new faster classloader that uses less memory for instance). Some people accept the startup time, others find it too annoying to use Java on the desktop. YMMV.
Sorry, but this is out of date. Slow start-up time is not an issue with current (1.5) VMs. Even full-featured Java applications like JEdit or Moneydance start up within a few seconds. This is no more annoying that any other typical KDE or GNOME application.
Way to go with great debating technique! Of course you could say 'show me evidence', and I could indicate years of data that back my case, but how much more effective to just post this word!
If I thought it was small possible economic problems and the propsed solution was likely to prevent a major problem, I would agree.
What I don't understand is how you so confidently assume that there isn't going to be a major problem. All you are implying is that the math of climate change is uncertain. Well you are right - it is! But the entire range of uncertainty still suggests warming with serious consequences. The models have a large error range, but all models indicate a major problem.
I am certainly willing to consider issues of dealing with rising seas. However, you have not demonstrated that the warming is caused by the CO2.
Only someone with little or no scientific understanding could come up with a statement like this. CO2 is a greenhouse gas. Nearly doubling its amount in the atmosphere has a warming effect that is well understood chemically. To assume that this won't have a warming effect is scientifically illiterate.
Beyond that, you consistently assume in the preceding paragraph and through the rest of this post that we CAN do something signicant.
Yes - stop pumping out CO2 is one thing.
You and many others need to contemplate and read and understand the difference between what is the "right" path and what is actually possible. It is very much the difference between the mathematician and the engineer. I am an engineer.
No. You and a few others need to understand the difference between what must be done and your limited view of what you think we are capable of doing. No amount of engineering skills can change the results of the math.
This sounds like wishful thinking to me. Sure, you can find some people who will say so (does this reasoning seem familiar?), but not many, and there are plenty of well respected economists who vehemently disagree.
This is totally irrelevant, as are your points about other countries reducing hydrocarbon use. Hydrocarbon resources are going to run out (or rather, it is going to become economically un-viable to continue using them at current rates). There will need to be a major reduction in their use anyway. We have a choice - do it now in a controlled fashion, or let our economies collapse later.
Let me see if I understand this. We have never managed climate before, but we are going to do this gigantic effort based on some computer models which cannot be tested against real world conditions (I choose not to get further into model esoterica).
And in doing so you neglect that the computer models are, of course, tested against real world conditions by attempts to model past climate.
Let us assume that the chances that the models are 90% correct (and I would be surprised if you would give it that high a probability).
Depends on what you mean by correct. Global warming is an established fact, not matter what one can argue is causing it.
Applying these predictions to dramatic policy changes is both arrogant and supremely foolish.
It is arrogant and foolish to note that the seas are rising (they are) and we need to think about how to prevent widespread damage? On the contrary, to do nothing in the face of established facts is foolish.
Certainly economics is not a zero-sum situation, but rapidly depreciating huge productive facilities and replacing them (especially with solutions far from optimized or even proven) is going to require enormous amounts of resources diverted from somewhere else - and that is quite simply going to hurt, and in this case, hurt very badly.
Well, big deal! The consequences of not doing this are going to hurt.
But the terrible consequences that large numbers (your metric) of well reqpected economists do agree upon are not to be considered because "we have never don it before."
I'm sorry, but I think this inconsistency in standards is simply not very convincing.
What inconsistency?
We are already 'managing climate' by pumping gigatonnes of CO2 into the air.
The consequences of not dealing with this are way beyond any petty considerations of economists. We are talking about the potential for hundreds of millions of deaths, not changes in unemployment, or gross national product.
Get a sense of proportion! I find the attitude that we should do nothing simply because of a lack of certainty to be an abdication of responsibility of a scale that is unbelievable. We know that the Earth is warming. There is no doubt. We know that CO2 assists with this. There is no doubt about this. We know that hydrocarbon resources are going to run out, assuming this hasn't already started (and there are signs that it may have).
To sit back and do nothing because of 'possible economic problems' is outrageous.
Mankind will survive one way or another. We have an enormous challenge that we can potentially turn into a positive enterprise.
Because the same modelling techniques work on different systems.
Since the climate is the time integral of these short term events, and the models are using the same physical laws, the predictions are either going to be parameterized out the kazoo, which certainly affects their credibility, or they should be sensitive to weather issues.
Climate is the time integral of short term events, but individual short term events are not influenced much by long term causes. Short term events tend to average out, so there is simply no point in trying to precisely predict millions of short term events to try and work out the long term ones.
Also, the long term events aren't caused by the short term ones - when you look at longer timescales, different factors become significant. You don't need to look at changes multi-decade shifts in water currents in order to predict the weather in a few days. You do to predict the climate in 100 years.
Weather makes climate and climate makes weather.
No, and this the fundamental mistake you are making. Weather does not 'make' climate. The short-term factors that influence day-to-day weather are largely irrelevant on the timescale of climate.
Think of fluid flow: You may get lots of small vortices, but these average. When you have a certain volume flowing into another, you will get the same volume flowing out, no matter what the complexity in between.
And I am saying that when it comes to policy, listening to "the majority of researchers" on a subject that is very far from settled is, in this case, simply insufficient. And if we had the real researchers on this forum, they could duke it out.
But they are duking it out, and in the appropriate forums - refereed journals and conferences. And they have come to a general (although not unanimous) concensus.
The difficulty of the social engineering required to take "precautionary" action dwarfs any effort mankind has ever undertaken. The probability of success is almost nil.
Why? The social engineering is equivalent to the amount of effort involved in a war. We have managed that sort of effort throughout the history of mankind.
But history and economics tells us that the effects of taking those actions will be non-compliance, wars, dramatic economic dislocations and large numbers of human deaths as a trickle-down consequence of the dramatic depression of first world economies.
We have no evidence at all of the results of what we need to do, as we have not done it before. There is a considerable opinion that reducing usage on hydrocarbons for fuel, for example, will hugely boost first world economies.
Although you did not say that, I will agree with what you are now saying about minority opinions. I choose to listen to the minority opinions of the people I mentioned because not only do I know them as human beings, I have had a chance to observe their methods and their level of care and self critique.
And I respect that, but it doesn't matter. Because a far greater number of people who are well qualified and good scientists have come to different opinions.
I think that land biomass sequesters a lot more carbon than you give it credit for. After all, when plants decay, the majority of their mass does not become CO2. It ends up as organic detritus on the forest floor.
Which decays.
If the biotic theory of petroleum is correct (and I am not suggesting it isn't), then where else would all the oil have come from in the first place?
Some of that organic detritus, but it is a totally insignificant amount compared to the mass of carbon trapped by precipitation of ocean plants over the history of the Earth.
How on Earth do you think a technological society can be maintained without it's associated energy infrastructure?
Who says our energy infrastructure has to be based on petroleum? Most of France's is nuclear, for example.
Life *will* become chaos for hundreds of millions of people (if not billions) on this planet. We will probably see the population return to, at most, pre-1850 or pre-1900 levels. Technology won't disappear overnight, but without the ability to replace equipment that has reached its end of life and without the people to replace the operators at their end of life, it will cease to hold the same amount of relevancy in our societies.
Woah there! There was a sudden break in logic. There is no reason why a reduction in population, or even a scarcity of hydrocarbons would mean that technology would start to disappear. In fact, the opposite is likely - crises drive innovation, and a reduced population would find itself in less trouble with a reduced hydrocarbon supply.
So you are saying that mechanisms that drive the daily weather are totally different than what drives the long term climate?
Yes.
On what KNOWN obervations do you base such an assertion?
Factors such as solar radiation, ocean current cycles.
Why should the same hitherto unknown factors that control the daily weather, also not control the weather over many days?
The factors that control the daily weather aren't unknown, they simply operate over smaller scales. They do control the weather over many days, but we are not interested in weather; we are interested in climate. The factors that control weather determine, for example, the amount of rain in New York on Tuesday. But in terms of climate, we don't want to know that. We want to know about average annual rainfall, so we deal with factors at larger scales and longer time intervals.
It is like the study of fluid flow. We may be interested in small-scale activity (vortices at fine scales) or large-scale activity (bulk flow).
Where is the cut-off time between the daily, weekly, monthly, yearly and ever longer peroiods of time where the laws by which the weather operates suddenly become known and predictable?
There is no cut-off, but the weather is controlled by limiting factors which allow averaging (for example, you may get a series of unusually hot summer days that are hard to predict, but you know that eventually autumn will come and the days will get cooler).
If short term weather is chaotic, by what mechanism is long term weather prevented from being so also?
It is not prevented from being chaotic, but it is restricted within certain bounds by things like the limited amount of solar radiation and feedback mechanisms.
"The Arctic ice is already 1/2 of the thickness it was decades ago. The permafrost in Northern Europe is melting. Glacier movements in Greenland have sped up many-fold."
Your point? This sounds a whole lot more like fear mongering than science.
No - these are facts.
Sounds like science suddenly got democratic.
It is. It is voted on by evidence, not minority opinion.
the question is: why these dramatic effects when the earth (according to paleo data) is not nearly as warm as it has been in the past, and (assuming your very questionable statement about midieval warm period temperatures) the current warming is very small?
Because when the Earth was much warmer in the past there were no global ice caps. We are not talking yet about much warming - we are talking about thermal balances. If you heat a bucket of ice it will hardly warm at all until it has melted. The decrease in polar ice shows that there is a lot of additional heat.
"The modeling of climate over periods of decades and centuries has nothing to do with the chaos of weather over days."
Would you care to prove this?
You link does a good job of doing this - they use different timesteps and scales to deal with climate forecasting than to do with weather forecasting.
Can you show that a system as with as many complex feedback mechanisms as earth is not chaotic on a particular time scale?
No. But that was not what I was saying. What I was saying is that the inability to predict detailed climate over a few days is irrelevant to the ability (or not) to predict climate over longer scales. I am certainly not saying there is no chaos.
The page is a project to create an extremely large ensemble for forecasting.
Yes, but this is a poor example, and you are confusing things.
As they say in their experimental strategy, what they are doing is similar to weather forecasting. However, it is not the same. They use longer timesteps for climate forecast, ignoring chaotic features of daily weather. So, the chaos of day-scale weather forecasts is irrelevant, which is what I said.
If we were in the '70s, the effects you mentioned would have been those supporting the *minority* scientist, who by your theory couldn't be right.
I did not say this. I did not say that all minority opinions are wrong. I said that minority opinions are generally wrong. Of course, some minority opinions are right, and become the foundation of future theory. But to assume that any minority opinion is worth listening to is just silly.
If you have a way to pick the minority opinions that are right, I would be interested. Simply picking the one that you would like to be right is not a good way.
I'm sorry, but when it comes to climate (as opposed to what area you have or do real since in), and when it comes to the methods of science *over time*, your statements ain't too hot.
Sorry, but I *am* an expert in numerical simulation, modelling and statistics.
There are many other systems involved that are not well understood. If we don't understand these major processes, we cannot model them.
No - we can simply do our best, as in any area of science.
Look - I am not saying you are wildly wrong, or what you are posting is rubbish. What I am saying is that the majority of researchers in the field disagree with you, so I am afraid it makes sense to listen to them and not you. Call it the precautionary principle.
Thus, according to the theory, if we heat up the Earth too much it might "suddenly" flip to a different equilibrium with a different atmosphere.
If the dinosaur-killer asteroid collisoin (with an energy equal to a million atomic bombs), combined with the Deccan Traps lava flows (around the same time) which pumped vast amounts of CO2 into the air, could not flip the equilibrium, then there is no way we can.
One can argue about 'scientific credibility' any time people wander far out of their area of expertise and into psuedo-religion. Which is what Lovelock has done.
I don't support all his Gaia ideas (by a long way!), but this is certainly his area of expertise - he has been working on climate regulation for decades.
No matter the language he uses, he has to be taken seriously, even if you are going to argue against him.
Slashdot Mirror
If we had the wherewithal to move to a different (habitable) planet, I suspect that we could also deflect asteroids, break up tsunamis, etc.
We have the capability of colonising other areas of the Solar System right now. A mission to the right sort of nearby resource-rich asteroid could 'bootstrap' our exploration and colonisation of the rest of the Solar System and beyond.
Defecting large asteroids may be possible in the near future, if we have enough warning, but the idea that we could break up tsunamis or prevent thousands of cubic kilometres of lava erupting from a supervolcano is absurd - it would require energies and technologies beyond anything we can currently conceive of.
Here's a clue to the solar system: There's nothing there within our reach that'll save us when we finally fuck up the planet irreparably. Get over the concept that we could move and start trying to fix the Earth first.
Actually, we could move pretty easily if we really wanted to; we are planning to return to the Moon in the near future, and hopefully some sort of base will be established at some point.
The problem is the the Earth isn't fixable. Even if we don't cause long-term (from our point of view) global warming, then eventually we are going to be hit by an asteroid, or there will be a supervolcano somewhere, or a mega-tsunami event. These would have effects way beyond anything we could achieve (the dinosaur-killer asteroid hit with the force of a million nuclear weapons!).
The idea that we could survive as we are long-term on this planet is totally unreasonable.
It's doing the same run-time cached-interpretation thing as python or anything else.
No it isn't.
Here is the full details of what happens when a Java program starts on a modern JVM.
Initially, a byte code interpreter starts running the program. This is reasonably fast (like Python), but nowhere near the performance of optimised native code. At the same time, a background thread is started which profiles the byte code and finds areas which are taking up the most time. These areas of the code are analysed in a lot of detail, and then aggressively optimised, with variable re-arrangements, method inlining, loop unrolling and (where possible) bounds checking removal. Then - and this is the key stage - the byte code is translated to highly optimised native code. When I say 'highly optimised' that is exactly what I mean. Amongst other things, the JVM can analyse processor pipelines and re-arrange the order of machine code instructions for optimal performance, and for most math operations, the JVM will embed processor instructions inline.
So, Java is not doing interpretation, at least not on any code that takes up time. It is running native code.
However, this takes time, which is why many benchmarks that run for just a few seconds don't show good performance for Java.
If you run a Java with the '-server' switch - which means do a lot of optimisation at the initial expense of response - then you will find that within a few seconds your program could well give close or equal performance to the output of many C compilers.
I think the difference from Python should now be clear.
Why does the performance of python need to be any worse than that of Java?
Because Java byte code is translated into high-performance native code, unlike Python.
A python script will have bytecode generated the first time and then run as high performance native code on any platform.
No interpreted byte code runs at a performance equivalent to native code on any platform. So say this is the case is simply wishful thinking. This is why a huge amount of effort is put into JIT compiling of these byte codes to native code. There are projects under way to get this working for Python.
I'm pretty sure a lot of the processing was direct python. And some languages make it easier to combine than others.
I'm sure it was, but if you read back, I was comparing Java with pure Python code. The point was that to get good performance the core parts had to be C. This dropping through to another language is inescapably messy.
There's no reason you couldn't do it in pure python - there's just no point.
Of course there is a reason you couldn't do this practically in pure Python - performance.
So they've written two copies of it because Java people refuse to reuse existing libraries?
No, they have written two copies of it because having to configure and re-build C source code on every platform on which you want to deploy is tedious and outdated. The Java implementation can be shipped as byte code and will then run as good performance native code on any platform.
My point is that python is just as suitable for image processing as Java, and usually more productive.
But Python isn't used for image processing! In the examples you give, Python is being used as a layer above C that does image processing. Python is a great language to wrap around C code for things like this, but to claim that Python is doing the processing is to mislead.
This is in stark contrast to Java. The standard Java API for image processing is JAI (Java Advanced Imaging). Although it can use C libraries if they are available it can (if these libraries or not present) also use a Java-only implementation which gives good performance.
No matter what the advantages of Python (and there are many) to claim that it is equivalent to Java in this area is simply incorrect.
If you want high-performance numerical code Java isn't that suitable either, you want C or fortran at a minimum and probably hand-coded assembler.
Have you researched Java's numerical performance recently? There is no need for C or Fortran. Java produces high-performance assembler that does a good job with numerical work.
But I've seen plenty of serious image processing done in python, remember the recent jigsaw solving story?
That is not serious image processing done in python. That is python using C. I used the phrase raw python. Any language can 'spoof' speed by calling out to a faster one.
I see the irony bypass was a complete success then.
Hey, I am British - we invented irony.
No need to be sorry, I'm glad to hear it. I'm a Java fan too. I'm stuck on Java 1.4 at the moment though, since we use it at work, so I haven't had time to test startup/Swing performance of the new versions. As soon as I'm finished with my math studies I'll install the recent versions at home and start developing. :-)
I don't know what they have done in Java 1.5 to speed things, but it really works - I am finding JEdit starting faster than typical equivalent KDE applications - like Kate (3 or 4 seconds).
Also, the look and feel of Swing has been changed - it was pretty awful in previous versions, but now has a sort of 3D shaded light blue look.
No-one seriously claims that. If productivity matters, why are you using Java rather than the 5-10x more productive Python?
Why are you assuming that Python is more productive?
I would love to see someone try and implement some of the code I have written (high-performance numerical work, image processing) in raw Python.
Best performing - SWT probably, but it is less portable.
SWT has recently had reports of performance problems on some platforms - Linux in particular. As Swing now uses hardware acceleration by default where possible (OpenGL,DirectX) the difference in performance between SWT and Swing is usually negligible. The fact that Swing is built-in to all JRE distributions gives it huge advantage.
Both still have the problem of JVM startup time though (another problem Sun is looking at, they are currently testing a new faster classloader that uses less memory for instance). Some people accept the startup time, others find it too annoying to use Java on the desktop. YMMV.
Sorry, but this is out of date. Slow start-up time is not an issue with current (1.5) VMs. Even full-featured Java applications like JEdit or Moneydance start up within a few seconds. This is no more annoying that any other typical KDE or GNOME application.
Liar.
Way to go with great debating technique! Of course you could say 'show me evidence', and I could indicate years of data that back my case, but how much more effective to just post this word!
If I thought it was small possible economic problems and the propsed solution was likely to prevent a major problem, I would agree.
What I don't understand is how you so confidently assume that there isn't going to be a major problem. All you are implying is that the math of climate change is uncertain. Well you are right - it is! But the entire range of uncertainty still suggests warming with serious consequences. The models have a large error range, but all models indicate a major problem.
I am certainly willing to consider issues of dealing with rising seas. However, you have not demonstrated that the warming is caused by the CO2.
Only someone with little or no scientific understanding could come up with a statement like this. CO2 is a greenhouse gas. Nearly doubling its amount in the atmosphere has a warming effect that is well understood chemically. To assume that this won't have a warming effect is scientifically illiterate.
Beyond that, you consistently assume in the preceding paragraph and through the rest of this post that we CAN do something signicant.
Yes - stop pumping out CO2 is one thing.
You and many others need to contemplate and
read and understand the difference between
what is the "right" path and what is actually possible. It is very much the difference between the mathematician and the engineer. I am an engineer.
No. You and a few others need to understand the difference between what must be done and your limited view of what you think we are capable of doing. No amount of engineering skills can change the results of the math.
But seriously, when are they going to deal with the myth that Java "is just as efficient as C++ these days"
Ah, but the MythBusters actually try things out and believe the evidence.
Slashdot myths this are famously immune to evidence, and therefore un-bustable.
This sounds like wishful thinking to me. Sure, you can find some people who will say so (does this reasoning seem familiar?), but not many, and there are plenty of well respected economists who vehemently disagree.
This is totally irrelevant, as are your points about other countries reducing hydrocarbon use. Hydrocarbon resources are going to run out (or rather, it is going to become economically un-viable to continue using them at current rates). There will need to be a major reduction in their use anyway. We have a choice - do it now in a controlled fashion, or let our economies collapse later.
Let me see if I understand this. We have never managed climate before, but we are going to do this gigantic effort based on some computer models which cannot be tested against real world conditions (I choose not to get further into model esoterica).
And in doing so you neglect that the computer models are, of course, tested against real world conditions by attempts to model past climate.
Let us assume that the chances that the models are 90% correct (and I would be surprised if you would give it that high a probability).
Depends on what you mean by correct. Global warming is an established fact, not matter what one can argue is causing it.
Applying these predictions to dramatic policy changes is both arrogant and supremely foolish.
It is arrogant and foolish to note that the seas are rising (they are) and we need to think about how to prevent widespread damage? On the contrary, to do nothing in the face of established facts is foolish.
Certainly economics is not a zero-sum situation, but rapidly depreciating huge productive facilities and replacing them (especially with solutions far from optimized or even proven) is going to require enormous amounts of resources diverted from somewhere else - and that is quite simply going to hurt, and in this case, hurt very badly.
Well, big deal! The consequences of not doing this are going to hurt.
But the terrible consequences that large numbers (your metric) of well reqpected economists do agree upon are not to be considered because "we have never don it before."
I'm sorry, but I think this inconsistency in standards is simply not very convincing.
What inconsistency?
We are already 'managing climate' by pumping gigatonnes of CO2 into the air.
The consequences of not dealing with this are way beyond any petty considerations of economists. We are talking about the potential for hundreds of millions of deaths, not changes in unemployment, or gross national product.
Get a sense of proportion! I find the attitude that we should do nothing simply because of a lack of certainty to be an abdication of responsibility of a scale that is unbelievable. We know that the Earth is warming. There is no doubt. We know that CO2 assists with this. There is no doubt about this. We know that hydrocarbon resources are going to run out, assuming this hasn't already started (and there are signs that it may have).
To sit back and do nothing because of 'possible economic problems' is outrageous.
Mankind will survive one way or another. We have an enormous challenge that we can potentially turn into a positive enterprise.
I really wish that users understood that when something goes wrong, it stays wrong, and won't be fixed by trying the same thing a dozen times,
Oh really? And why would that be?
Because the same modelling techniques work on different systems.
Since the climate is the time integral of these short term events, and the models are using the same physical laws, the predictions are either going to be parameterized out the kazoo, which certainly affects their credibility, or they should be sensitive to weather issues.
Climate is the time integral of short term events, but individual short term events are not influenced much by long term causes. Short term events tend to average out, so there is simply no point in trying to precisely predict millions of short term events to try and work out the long term ones.
Also, the long term events aren't caused by the short term ones - when you look at longer timescales, different factors become significant. You don't need to look at changes multi-decade shifts in water currents in order to predict the weather in a few days. You do to predict the climate in 100 years.
Weather makes climate and climate makes weather.
No, and this the fundamental mistake you are making. Weather does not 'make' climate. The short-term factors that influence day-to-day weather are largely irrelevant on the timescale of climate.
Think of fluid flow: You may get lots of small vortices, but these average. When you have a certain volume flowing into another, you will get the same volume flowing out, no matter what the complexity in between.
And I am saying that when it comes to policy, listening to "the majority of researchers" on a subject that is very far from settled is, in this case, simply insufficient. And if we had the real researchers on this forum, they could duke it out.
But they are duking it out, and in the appropriate forums - refereed journals and conferences. And they have come to a general (although not unanimous) concensus.
The difficulty of the social engineering required to take "precautionary" action dwarfs any effort mankind has ever undertaken. The probability of success is almost nil.
Why? The social engineering is equivalent to the amount of effort involved in a war. We have managed that sort of effort throughout the history of mankind.
But history and economics tells us that the effects of taking those actions will be non-compliance, wars, dramatic economic dislocations and large numbers of human deaths as a trickle-down consequence of the dramatic depression of first world economies.
We have no evidence at all of the results of what we need to do, as we have not done it before. There is a considerable opinion that reducing usage on hydrocarbons for fuel, for example, will hugely boost first world economies.
Although you did not say that, I will agree with what you are now saying about minority opinions. I choose to listen to the minority opinions of the people I mentioned because not only do I know them as human beings, I have had a chance to observe their methods and their level of care and self critique.
And I respect that, but it doesn't matter. Because a far greater number of people who are well qualified and good scientists have come to different opinions.
I think that land biomass sequesters a lot more carbon than you give it credit for. After all, when plants decay, the majority of their mass does not become CO2. It ends up as organic detritus on the forest floor.
Which decays.
If the biotic theory of petroleum is correct (and I am not suggesting it isn't), then where else would all the oil have come from in the first place?
Some of that organic detritus, but it is a totally insignificant amount compared to the mass of carbon trapped by precipitation of ocean plants over the history of the Earth.
How on Earth do you think a technological society can be maintained without it's associated energy infrastructure?
Who says our energy infrastructure has to be based on petroleum? Most of France's is nuclear, for example.
Life *will* become chaos for hundreds of millions of people (if not billions) on this planet. We will probably see the population return to, at most, pre-1850 or pre-1900 levels. Technology won't disappear overnight, but without the ability to replace equipment that has reached its end of life and without the people to replace the operators at their end of life, it will cease to hold the same amount of relevancy in our societies.
Woah there! There was a sudden break in logic. There is no reason why a reduction in population, or even a scarcity of hydrocarbons would mean that technology would start to disappear. In fact, the opposite is likely - crises drive innovation, and a reduced population would find itself in less trouble with a reduced hydrocarbon supply.
So you are saying that mechanisms that drive the daily weather are totally different than what drives the long term climate?
Yes.
On what KNOWN obervations do you base such an assertion?
Factors such as solar radiation, ocean current cycles.
Why should the same hitherto unknown factors that control the daily weather, also not control the weather over many days?
The factors that control the daily weather aren't unknown, they simply operate over smaller scales. They do control the weather over many days, but we are not interested in weather; we are interested in climate. The factors that control weather determine, for example, the amount of rain in New York on Tuesday. But in terms of climate, we don't want to know that. We want to know about average annual rainfall, so we deal with factors at larger scales and longer time intervals.
It is like the study of fluid flow. We may be interested in small-scale activity (vortices at fine scales) or large-scale activity (bulk flow).
Where is the cut-off time between the daily, weekly, monthly, yearly and ever longer peroiods of time where the laws by which the weather operates suddenly become known and predictable?
There is no cut-off, but the weather is controlled by limiting factors which allow averaging (for example, you may get a series of unusually hot summer days that are hard to predict, but you know that eventually autumn will come and the days will get cooler).
If short term weather is chaotic, by what mechanism is long term weather prevented from being so also?
It is not prevented from being chaotic, but it is restricted within certain bounds by things like the limited amount of solar radiation and feedback mechanisms.
"What matters in science is majority opinion ..."
"Science works by testing ideas against evidence, not as a popularity contest"
Make up your mind.
In science, majority opinion is based on whether or not ideas pass the test of evidence, not 'popularity'.
"The Arctic ice is already 1/2 of the thickness it was decades ago. The permafrost in Northern Europe is melting. Glacier movements in Greenland have sped up many-fold."
Your point? This sounds a whole lot more like fear mongering than science.
No - these are facts.
Sounds like science suddenly got democratic.
It is. It is voted on by evidence, not minority opinion.
the question is: why these dramatic effects when the earth (according to paleo data) is not nearly as warm as it has been in the past, and (assuming your very questionable statement about midieval warm period temperatures) the current warming is very small?
Because when the Earth was much warmer in the past there were no global ice caps. We are not talking yet about much warming - we are talking about thermal balances. If you heat a bucket of ice it will hardly warm at all until it has melted. The decrease in polar ice shows that there is a lot of additional heat.
"The modeling of climate over periods of decades and centuries has nothing to do with the chaos of weather over days."
Would you care to prove this?
You link does a good job of doing this - they use different timesteps and scales to deal with climate forecasting than to do with weather forecasting.
Can you show that a system as with as many complex feedback mechanisms as earth is not chaotic on a particular time scale?
No. But that was not what I was saying. What I was saying is that the inability to predict detailed climate over a few days is irrelevant to the ability (or not) to predict climate over longer scales. I am certainly not saying there is no chaos.
The page is a project to create an extremely large ensemble for forecasting.
Yes, but this is a poor example, and you are confusing things.
As they say in their experimental strategy, what they are doing is similar to weather forecasting. However, it is not the same. They use longer timesteps for climate forecast, ignoring chaotic features of daily weather. So, the chaos of day-scale weather forecasts is irrelevant, which is what I said.
If we were in the '70s, the effects you mentioned would have been those supporting the *minority* scientist, who by your theory couldn't be right.
I did not say this. I did not say that all minority opinions are wrong. I said that minority opinions are generally wrong. Of course, some minority opinions are right, and become the foundation of future theory. But to assume that any minority opinion is worth listening to is just silly.
If you have a way to pick the minority opinions that are right, I would be interested. Simply picking the one that you would like to be right is not a good way.
I'm sorry, but when it comes to climate (as opposed to what area you have or do real since in), and when it comes to the methods of science *over time*, your statements ain't too hot.
Sorry, but I *am* an expert in numerical simulation, modelling and statistics.
There are many other systems involved that are not well understood. If we don't understand these major processes, we cannot model them.
No - we can simply do our best, as in any area of science.
Look - I am not saying you are wildly wrong, or what you are posting is rubbish. What I am saying is that the majority of researchers in the field disagree with you, so I am afraid it makes sense to listen to them and not you. Call it the precautionary principle.
Thus, according to the theory, if we heat up the Earth too much it might "suddenly" flip to a different equilibrium with a different atmosphere.
If the dinosaur-killer asteroid collisoin (with an energy equal to a million atomic bombs), combined with the Deccan Traps lava flows (around the same time) which pumped vast amounts of CO2 into the air, could not flip the equilibrium, then there is no way we can.
One can argue about 'scientific credibility' any time people wander far out of their area of expertise and into psuedo-religion. Which is what Lovelock has done.
I don't support all his Gaia ideas (by a long way!), but this is certainly his area of expertise - he has been working on climate regulation for decades.
No matter the language he uses, he has to be taken seriously, even if you are going to argue against him.