Sunday, December 19, 2010

The Global Warming Saga: A simple view on an over-complicated story.

Andrew D Atkin:

I have already written on this issue, but I think I can make a clear and simple 'master picture' which I would like to contribute.

As follows:

Nobody is arguing that an increase in CO2 will lead to some increase in average global temperatures, in itself. But it is only a very mild greenhouse gas. If we doubled the CO2 concentration in our atmosphere (from where it is now) we would get around a 0.5 to 1 degree Celsius increase in average global temperatures, which is benign and probably even eco-beneficial.

[Our world would support more plant growth overall if it were a little warmer than today, and CO2 is plant food. CO2 is literally (and vitally) an atmospheric fertiliser].

So where is the argument? How the apocalypse? The argument from the IPCC (Inter-Governmental panel on Climate Change) is based on the idea that the marginal increase in warming coming from an increase in CO2 will be amplified by positive-feedback; that is, a slight increase in warming will heavily snowball via reactive processes in the atmosphere, so that we end up with a temperature increase not just about a half a degree greater (from the CO2) but eventually more like 3 or 5 degrees greater (or more?).

If I may speak frankly...Bollocks! If it were true that the earth responds to marginal increases in temperature with a positive-feedback loop (or "snowball" for common language), then we would have already turned into a Venus countless times over from temperature variations that have always (and continuously) happened throughout our ecological history (variations in the suns output being maybe the most significant).

Small, common changes in temperature do not lead to major changes in climate. The medieval warm-period did not turn into a runaway greenhouse effect.

So where did the IPCC get its evidence from? How did it reach its 'weird' conclusions? It did so from computer models. They developed computer programmes that (they claimed) simulated the environment.

Straight off the mark we should be laughing. Because to even suggest you can *accurately* simulate the climate for over a long period of time is a nonsense. This is like claiming divine knowledge - that we don't have. To model the environment so as to simulate it you have to understand it explicitly and be able to measure it - with all the many, many variables accurately accounted for.

First, we just can't do that (yet). We can barely measure known climatic variables, let alone (fully) understand how they all work against each other (which is no doubt why there is still so much debate, research and theory within climate science today). And even if we could measure and account for all the climatic variables (so as to develop accurate long-range cybernetic simulators) we would still require enormous long-term testing so as to develop a simulator that has a reasonable accuracy.

To believe that the IPCC has a "from the God's" cybernetic crystal-ball is one hell of a stretch.

Addition: 30-7-11: Here is an article that appears to vindicate my assertions. You will notice they're talking real world data - not computer models. The former answers to the latter, not the other way around. Models are the guess - real world measurements are the fact.


My conclusion is that we cannot take the computer models seriously. And also, the simple fact that trivial temperature variations (that happen all the time) do not lead to runaway-greenhouse effects suggests the obvious: that the IPCC is a political organisation established to achieve a political end. And an end that has nothing (nor ever did have anything) to do with saving the world from global warming.

Here is a relevant article by Lord Christopher Monckton, probably much more relevant than what I have just written:

And finally, people like Richard Lindzen (people who study direct measurements of climate - not people who pretend to believe in garbage-in/garbage-out computer simulators) have (supposedly) discovered that the climate is in fact a negative-feedback system (like a natural governor that tends to keep things constant). The latter is far more believable to me, because it would explain the relative stability of our climate in response to normal temperature variations.

...Monckton did another good talk, relating to my central points (posted 11-05-14):

Monday, December 13, 2010

Congestion charging? Yes!

Andrew D Atkin

The principle:

Let's say you started up a coffee shop in town and it proved to be super popular, because unlike the rest of New Zealand you understood the importance of quality-control when it comes to food. So, likewise, you then end up with more demand than you can supply. So what do you do?

You can either actively control (reduce) the demand by pushing up your prices, or you can let the demand manage itself (passively) by allowing big queues to form out your door (so with the queue your customers pay not with more cash, but with their time). Those are your two choices.

From a macro-economic outlook, using queues to control demand is wasteful; because, of course, "human resources" are left standing idle when they could have otherwise been producing (or whatever).

The economy will be richer overall if your coffee shop just increases its price to control the demand. Indeed, I believe that queues, at base, are just an expression of miss-pricing.


It's funny how people blame roads for congestion as though they are the problem in themselves. To me, that's a bit like blaming a coffee shop for creating a queue--So get rid of the good coffee shop?

Road congestion is basically just another queue - a queue of cars lining-up to use a popular road. It's a particularly expensive queue because it wastes a heck of a lot of fuel and wears out your car, as well as wasting your time.

Road congestion exists, of course, because it can be too difficult or expensive to build enough supply (more roads) and/or because the public will not accept demand-control for roads via pricing (aka congestion-charging).

Well, this is something that we urgently need to revise, especially for in a city like Auckland.

As follows:

The days of stopping for cumbersome toll-gates are gone. We can precisely and adaptively (adaptively = rapidly changing prices to suit immediate demand) price the roads anywhere and at any time.

Cameras can be set up to read your licence plate when you enter and exist a toll-road, and the administrative process can be fully automated. You can even option for direct-debit from your bank account and have the toll-bill sent to you via email.

We can go further than this still. If we employ a passive RF-tag on all cars integrated with a GPS system then we can abandon the cameras altogether, and people can review costs for going anywhere via the internet, using a real-time 'toll map'.

Sound expensive? Well, the money collected from tolling doesn't just disappear - it will be used to pay your rates. What the council doesn't get from tolls it will get from other means. So, congestion-charging isn't so much another cost, it's basically just demand-control. And you will of course save a lot of money from not having to idle your car nor waste your time in that queue.

The reaction:

If roads are tolled adequately enough so as to get rid of congestion, then the less affluent will tend be more careful with how they drive (or not). You will see an increase in demand for buses, car-pooling, and integrated trips (during peak demand time, at least). You will also see more telecommuting.

But travel will be much easier and more rapid alround, products will be cheaper to buy (because you don't pay more via the mark-up to cover excessive freight costs), and although you will get a toll-bill it will be met with a relative reduction in rates. Again, the money does not just disappear.

You should also get a notable increase in economic productivity, not only from the reduced (total) transport costs, but because un-congested roads can greatly improve accessibility between businesses and customers, providing for some improved economic integration*.

*Note: Some researchers claim that greatly increasing urban density will achieve this effect. I doubt it. The reader might like to review Phil McDermott's piece on this issue:

Congestion-charging Versus Tolling:

By my definition, congestion-charging is where you only toll enough so as to control congestion. Contrasting, straight tolling is based on maximising revenue.

Where congestion-charging is employed you will often see no toll during the off-peak times (because you just don't need it because there's no congestion to battle anyway), whereas there will always be a toll where revenue-based tolling is employed.

I prefer congestion-charging because it provides for the full utilisation of the road, and therefore maximises its productivity. With revenue-based tolling you end up pushing people off the road for even when there is a surplus of capacity.

Political resistance:

It was my prediction from a few years back that the ARC would resist investment in congestion-charging. Why? Because they have their Smart Growth vision and it just won't go away. And that is a vision of a high-density Auckland with concentrated development built up around the railway line...

The (then) ARC knows as well as I do that trains can't compete (on any level) with buses operating on congestion-free roads; and so they would have huge difficulty (as they do already) justifying wasting billions of dollars on their Smart Growth/rail visions ahead of the bus (and other) options, if rational congestion-charging were to be employed. So they want roads sick - not healthy.

And so far my prediction seems to be holding up. Alas, visionaries are simple thinkers: Something is either consistent (and therefore 'good') or inconsistent (and therefore 'bad') with their end-goal visions. Period.

The other resistance is trying to sell congestion-charging to the public who tend to only see an extra bill being presented to them. What can I say?
People, it's just so silly to use queues to control demand. Ultimately, you only end up paying more!


Addition: 14-2-11:

Public-Private Partnerships?

I am suspicious about the use of public-private partnerships (PPP's) in roading investment, because I don't know how it may affect the (toll-based) business model.

Will a government operating under a PPP be forced to operate a revenue-based model, and therefore employ traditional tolling rather than congestion-charging? If so, then it might be best to just stick to public funding for roads, at least for roads that are essential for the national interest. (And if the roads are in fact essential, then that would mean there is no effective investment-risk, and so the government would in turn have no good reason to have the private sector as a part-owner share what risk?)

As I explained earlier, revenue-based tolling reduces productivity because it reduces the utilisation of the road. Though the revenue from the toll-road would be increased, the productivity of the road will be reduced. And productivity is what we're after - not a money-go-round.

The private sector is of course exclusively focused on their bottom-line. The [non-monetised] social value of their operation means nothing, and because of this they can sometimes be inefficient and wasteful with respect to an ultimate social advantage. When this is the case, there can be a good argument for direct and maybe exclusive government involvement. Roads may be an example of this.


Addition: 17-4-14:

Letter to Julie-Anne Genter:

The included letter was sent 09-01-12. I thought I would include the idea for interest as it's relevant to this post:

Green Party MP
Julie-Anne Genter

Hi again,

I had a look through the idea of removing minimum parking requirements, and I agree with it. In my view the only reason why we had them in the first place is probably because we just never (previously) had the tools to meter all (or nearly all) parks, meaning we haven't yet been able operate a rational market model [addition: The same can be said for most roads - we never charged for them because it was just too impractical, except in exceptional situations such as for costly new bridges].

However, it's clear to me that we now have the tools to toll any park or road economically, using the system I suggest.

I wanted to forward this to you because it has become obvious to me that this system, surely, is by far the best way to go about it. And also it would need to be a national initiative for if it were to ever go ahead.

As follows:

1. Mandate passive RF-chips on all cars in New Zealand, to be fixed onto license plates for when the cars get their WOF. This of course provides an electronic signature for all cars registration.

Passive RF-chips are so cheap in themselves that they can be considered costless.

2. From here, you can install an RF-reader embedded-in or placed on top of the road. The RF-reader would basically be an extremely crude cellphone-type device that records the registration of all cars that pass over it. From here it can send a text via wireless internet to inform a master server of what/when/where a given car went through the gate. The reader can be solar-powered (only a tiny amount of power would be required to run it). It would likewise accumulate data and maybe send a text to a master server with its records, once a day. This is extremely simple and easy to install technology - no wiring required.

3. From here, your server will have all the information it needs to bill a driver for both toll roads and parking. Every driver will have an established account, and people can be sent a bill for their usage, usually as a PDF-file every month, and pay online too.

The administration can be (and naturally would be) almost totally automated.

4. Private sector ownership of car parks and roads will have their revenue paid to them through the Ministry of Transport, from the MoT's master server. Naturally it needs to be based on one national system to be practical.

-No one will want to muck about with multiple bills from multiple servers, and nor do they need to if you get the system right from the beginning and build a single core-system as the base. This is also why it would need to be an initiative developed by central government. It needs to be a national system so all cars can be charged. It's a bit of a chicken and egg thing - all the cars need to be chipped first.


I cannot conceive of a more efficient and practical system for providing user-charges on roads. And to stress there is no fundamental reason for it to be unduly expensive - the supporting technology is inherently cheap.

I do not believe the proposed system exists as of yet (from what I know at least?), but the government could certainly commission its development. There is no question it would work. It's based on nothing more than a crude information exchange using well established technologies. It's all solid-state electronics and therefore inherently reliable and low maintenance.

As I see it, it gives us the foundation to economically rationalise road usage which in turn allows us to do away with minimum parking requirements. Parking can become just another component of the market economic model.


Another thing you can do with this system is install the equivalent of speed camera's, for cheap. It's just a matter of embedding two readers, say 50 meters apart, on any given stretch of road so it can likewise measure vehicle speed as it enters and exists the gates, plus details.

I wonder how this would affect the road toll, having a "speed trap" on maybe every dangerous corner?

Road management:

The RF-readers can inform us of traffic conditions in real time, and very accurately. This can obviously help with traffic management.

Also, we can have congestion-charging with this system and on a detailed level, using maybe many toll gates because they're so cheap to install, anywhere. People can use the internet to get a detailed perspective on travel/parking costs at any given time of the day, as rates are always displayed online.

Reducing congestion is by far the most significant way we can reduce carbon emissions from road transport. Stop-and-go operation is the great "evil" of transport inefficiency in an urban environment.


Having a detailed time/location record of people's cars, that can be accessed when required, can no doubt help to fight crime.


I hope you found this idea of interest.

Thanks for your attention,

Andrew Atkin