Saturday, August 31, 2013

The State of Oil

Former Saudi oil minister Sheikh Yamani is reputed to have said
The stone age didn’t end because we ran out stones.
By this I guess he meant that the industry should watch its success at maintaining a price cartel, because that would lead to intensified research on alternatives. Since the oil crises of the 1970s and 1980s, OPEC has tried to keep pumping oil at the rate the market demands, to avoid the kind of price shocks that lead to renewed electric car research, better public transport, and so on.

Oil prices adjusted to 2013 dollars. Source:
How successful have they been? Prices fell rapidly since the last major conflict-induced price shock in 1979, the Iranian revolution. While the Gulf War of 1990 caused a price spike, this was nothing on the scale of the previous price shocks. The Iraq war in 2003 is not a serious candidate for the ongoing increase in prices, since this has continued well past the point where Iraqi oil production resumed. The Arab Spring movement of 2011 can’t be blamed for a major decline in output either, as the countries most affected are not major oil producers. Libya, for example, is only responsible for about 2% of worldwide oil production, though it’s 17th in the world in total production. And Libyan oil output us more or less back where it was before the 2011 uprising.  Remember also that there has been a worldwide economic crisis since 2007, with many countries battling to recover since. So the oil price spike hasn’t been driven by major conflict or an upsurge in demand. The worst of the 2007–2008 global crisis only pulled prices back a little, bringing them nowhere near the inflation-corrected stable average of about $20 (corrected to 2013 US$).

It is plausible that this price spike is being driven by growing difficulties in extracting oil fast enough to meet demand. If you don’t know about peak oil theory, this may be a good time to find out.

Saturday, August 24, 2013

Climate Change Lines of Evidence

Much of the contrarian position on climate change relies on question the sensitivity of the climate to varying CO2 content of the atmosphere. The mainstream science says that doubling the CO2 concentration should add about 3°C, with some variation depending on the starting conditions (e.g., the amount of ice at the start). Contrarians claim that climate sensitivity to increased CO2 is much lower, perhaps at most 1°C per doubling – and some even claim the greenhouse effect is already saturated, meaning more CO2 will have no effect. All these contrarian claims are simply claims without evidence to back them up.

The evidence I’ve found so far strongly suggests the contrarians are wrong. Which is a real shame, because the contrarian argument is winning as long as there is no strong worldwide move to cleaner energy. Since I don’t want disaster, I would really like them to be right. Unfortunately, nature has no concern for our preferences.

One of the problems with understanding why adding CO2 to the atmosphere drives climate change is there isn’t a simple equation that shows how the physics works. To calculate the actual effect requires taking into account the overall complexity of the planet, including the nature of the atmosphere that not only varies in density with altitude but also composition – water vapour for instance is not only highly variable in its geographic spread but the range of altitudes over which it is present.

If you want a comprehensive understanding of the greenhouse effect and how it applies to a whole planet, you have little option short of reading a textbook (e.g. RT Pierrehumbert’s Principles of Planetary Climate).

Nonetheless there are plenty of lines of evidence that make the case even if you don’t have a deep understanding of physics and calculus. Here are some.

Faint Young Sun

4-billion years ago, the sun’s intensity was only 70% of its current energy output. That would not have been sufficient under today’s conditions for the oceans to be liquid, yet the geological evidence is that there were liquid oceans that long ago. During that time, there was no life as we know it, and greenhouse gas concentrations would have been a lot higher than they are now. Greenhouse gases are the best explanation for liquid oceans at that time, though there is a lot of uncertainty given how long ago this was, and the limited range of geological data. Even so, we can be reasonably sure that a CO2 level ten times or more the current level would have had a very significant warming effect, and that coupled with other greenhouse gases could have made the planet warm enough for liquid water.

Snowball Earth

Several times in the distant past, with the last such event about 630-million years ago, the Earth froze over completely, almost to the equator. Under these conditions, CO2 cannot be drawn down from the atmosphere. There is unlikely to be sufficient plant life – even phytoplankton – able to consume CO2, and natural chemical processes that draw down CO2 require liquid water. In a snowball earth, CO2 in the atmosphere gradually increases until there is enough of a greenhouse effect to start melting the ice. This CO2 builds up from emissions from volcanoes, normally a minor effect offset by weathering, where rocks high in calcium dissolve in carbonic acid (CO2 in water) to form carbonate rocks such as limestone. Weathering requires liquid water, so the chemistry required stops if the planet ices over completely. Consequently any CO2 vented by volcanoes that escapes the ice cover into the atmosphere stays there. Without a greenhouse effect, the ice would not melt because ice has a high albedo (fraction of incident light reflected), and any region covered in ice reflects a high fraction of incoming solar energy back to space. Absent an increased greenhouse effect, a planet can only shed such a complete ice cover after a wait of millions – possibly billions – of years for the sun to become warmer.

A snowball earth can eventually end in a return to a warmer world as a result of the steady accumulation of CO2 in the atmosphere, leading to sufficient warming to melt the ice back to the polar regions – some simulations show this taking as little as 2000 years, a sharp contrast to waiting billions of years for the sun to warm up.

Milankovitch Cycles and Climate. The top three graphs
derive orbital parameters and the fourth (from the top)
shows resulting variation in northern hemisphere solar
energy. The bottom two graphs are indications of past
climate variability that can be related to the fourth
graph. For more detail, see WikiPedia.

Milankovitch Cycles

Serbian scientist Milutin Milanković, through laborious calculations, demonstrated that variations in the Earth’s orbit over millions of years corresponded to movement in and out of ice ages. Physics calculations of the change in incoming solar energy cannot account for the temperature swings needed to change between states of kilometres-thick ice caps and no ice over continental-scale regions. On the other hand, we know that the oceans’ capacity for dissolving CO2 is temperature dependent. As temperatures increase, the oceans release CO2. This increases the greenhouse effect, and hence amplifies a temperature increase. The opposite applies when temperatures are decreasing.

In this scenario, CO2 is acting as a positive feedback, an amplifying effect, rather than as a primary driver of climate change.

Reverse the scenario: if a change in the Earth’s orbital parameters causes cooling, the oceans dissolve more CO2 and once again CO2 amplifies the change.

Other feedbacks include a change in ice (less ice means less heat is reflected to space) and increased water vapour (a strong greenhouse gas) but these effect are insufficient to explain the temperature swings.

Mass Extinctions

PETM in context. In geological time the PETM event is for
practical purposes instantaneous – yet happened about 50
times slower than current climate change. Source: WikiPedia.
Our planet has undergone several mass extinctions where a large fraction of life was eliminated. These mass extinction events generally follow a massive planetary climate change. In some such events, the change has been a rapid temperature rise, preceded by a rapid increase in greenhouse gases. Good examples are the end-Permian extinction (the biggest of the mass extinction events, about 253-million years ago) and in the Paleocene–Eocene Thermal Maximum (PETM). The PETM is the more recent of these two events, and hence a bit easier to study in detail. The evidence if that temperatures increased by 6°C over about 20,000 years. Such an increase could be caused by a quadrupling of CO2. It’s sobering to consider that such an increase over 20,000 years caused a major extinction event, with an average increase of 0.003°C per decade, compared with the current rate of increase of about 50 times faster.

Other Planets

Planetary temperatures. Note that though Mercury
is much nearer the sun than Venus, Venus has a much
higher average surface temperature.
 Source: NASA.
The temperature of a planet is dependent on three major variables (with other factors like oceans contributing):

  • distance from the sun (or in effect the amount of incoming solar energy)
  • albedo (the fraction of incoming energy that is reflected)
  • atmosphere (ability of the atmosphere either to reflect energy back to space, or to slow the rate of energy flow out to space)
The planets from the sun outwards should be increasingly warm, if the first effect is the only one. This is mostly the case, but Mercury has a lower average surface temperature than Venus, despite being much closer to the sun (57,910,000 km, versus Venus at 108,200,000 km). Much modern climate science derives from study of the temperatures of other planets. Venus has a dense atmosphere mostly containing CO2, so it has an extreme greenhouse effect. Other planets have different atmospheres (Mars’s atmosphere, for example, is also mostly CO2 but much less dense than the Earth’s) , and modelling these variations is a starting point in the theories that today are used to model the effect of increasing CO2 on our planet’s climate.


All these lines of evidence are a small part of the picture. If here is indeed a very low climate senstitivity to increased CO2, all of the theory explaining these events would be wrong. So contrarians not only have to explain why prediction of future climate change are wrong, but also have a reasonable alternative explanation for all these lines of evidence.

The current theory of climate, which takes into account variations in solar output, changes in the atmosphere and various modes of redistribution of energy around the planet, has been well tested in several ways, including modelling the paleoclimate, forecasting climate change then checking as new data comes in, and hindcasting (testing predictions against known data). In addition, there are other lines of evidence like shifts in the range of temperature-dependent plants and animals.

The current theory of climate started from understanding the shifts between ice ages and temperature on other planets of our solar system. The theory continues to be refined as more data becomes available and more computational power is available for more complete models. So far, the major effect of improved models has been refinement rather than refutation. Every now and then a contrary result appears, but it does not stand up to rebuttal.

Every now and then, a model prediction turns out to be incorrect. It would be surprising were that not the case, with such a large-scale, complex system to model. Pointing at such flaws as indicating there is no problem and we should continue with business as usual is silly – like a morbidly obese patient deciding not to lose weight when a heart attack scare turns out to be a false alarm.

Without a strong greenhouse effect, the Earth could not have had liquid water when the sun was only emitting 70% of its current energy. Without a greenhouse effect, the Earth could not have escaped a snowball state. The greenhouse effect is required to amplify the effect of Milankovitch Cycles, otherwise orbital variations are insufficient to explain deep difference between glacial and interglacial climate. In all these cases, CO2 has to play a prominent role. There is no other greenhouse gas that can vary on a sufficient scale to make a difference. The kind of very rapid climate change – particularly warming events – that has triggered some of the biggest mass extinction events can also only be explained by the greenhouse effect. Finally, the greenhouse effect has been thoroughly studied for other planets and nothing else can explain the extremely high surface temperature of Venus.

Sunday, August 18, 2013

It’s the Energy

Bill Clinton’s campaign strategist James Carville coined the phrase The Economy Stupid, often since echoed as It’s the X, stupid as a mantra for keeping focused. I prefer not to call people stupid. So let me just say we should pay more attention to energy flows when we talk climate change, otherwise stupid or not, nothing will matter.

There is a certain rate of energy flow to the planet from the sun. Unless the planet can match that inflow by exactly the same outflow, the system is not in balance. The only way our planet can shed excess energy is by radiation to space. The only way we can increase the net outgoing radiation is by a temperature increase. So the thing of real concern is radiative energy balance [Pierrehumbert 2011].

So why then does everyone keep going on about temperature?

It makes sense to measure the temperature on the ground because it has to increase if energy is flowing in faster than energy is flowing out (energy flow in > energy flow out), until the rate of outgoing energy matches incoming.

The problem is that by focusing on temperature as the one measure of interest, we miss a rather vital point:
If the temperature is not increasing while net energy flow in exceeds net energy flow out, we are not shedding the excess energy.
Why is this important? To see why, we need to understand where the extra energy can go if it’s not increasing surface temperatures.

There are two primary ways the planet stores extra energy without raising surface temperatures: ocean sequestration and latent heat. Let’s look at these in more detail:
  • ocean sequestration – only the upper layer of the open is sufficiently transparent to infrared to radiate to space; if energy is transported deeper by currents, most of it stays there
  • latent heat – when a substance undergoes a phase change (solid to liquid or liquid to gas) a significantly higher amount of energy is needed for the phase change than for a temperature change of 1 degree (sensible heat is the term used for energy that results in a temperature change); here are some figures for water, all for 1kg (to a good approximation, a litre; we measure energy in units of a joule or J):
    • energy to increase temperature by 1°C: 4.2kJ
    • latent heat of melting: 334kJ (about 80 times the energy needed to raise the water by 1°C)
    • latent heat of evaporation: 2.3MJ (over 500 times the energy needed to raise the water by 1°C)
Latent heat vs. sensible heat. It takes almost as much
energy to melt 1kg of ice without changing the temperature
as to raise 1kg of water through 100°C; evaporating 1kg of
water takes 5 times as much energy as raising water by
It is instructive illustrate the relative amounts of energy needed for latent and sensible heat by comparing the energy needed to raise water by 100°C (the range over which it can be a liquid at sea level), with the energy needed to melt ice and the energy needed to evaporate water.

Ocean sequestration is important because many ocean ecosystems rely on the current stratification of ocean temperature, warm on top, icy-cold in the depths. Mess with that in a big way, and consequences for biodiversity and the food change could be dire, and are unpredictable with current knowledge. Also, temperature stratification is a major driver of ocean currents, something we disrupt at our peril. Another important factor is that this energy can periodically be released to the atmosphere resulting in sudden spikes in temperature (e.g. an El Niño event).

What are some consequences of the latent heat part of the equation?

If ice is melted, that is a positive feedback or amplifying effect. In other words, the radiative energy imbalance increases. Why? Because ice is highly reflective (in physics terminology, it has a high albedo). As long as the ice is only thinning there’s no change in albedo but once bare ground or water is exposed, the newly exposed darker surface absorbs more energy. If a region of the planet absorbs more energy, that pushes the net energy balance upwards, meaning temperatures have to increase further to balance the equation.

Latent heat of evaporation is not our friend either, for a different reason.

Latent heat is in general terms a big deal because it can use up so much more energy than temperature change while a phase reversal can release that energy as a temperature change. For example, if water vapour precipitates out of the air, the energy of vaporisation returns to the air as a temperature rise. Latent heat is a major driver of extreme weather [Liu et al. 2011] because of the massive differential between the energy needed for a phase change and the energy needed for a 1°C shift.

The really important thing about all this however is that these energy sinks can take up large amounts of incoming energy without changing the radiative energy balance. In other words, they do not fix the problem of energy flow in > energy flow out.

Why does this matter? Because an apparent slowdown in temperature increase is not good news if the energy imbalance is still there. The extra energy is having a number of effects most of which are not positive – and temperature increase is still going to happen.

If this is so important, why do we mostly talk temperatures not energy? Because temperature is something we can experience directly, something that is easy to measure and also a direct reflection of the energy increase, if not the only one. But most importantly, the surface temperature has to increase eventually so outgoing radiation can match the inflow. Why has energy not figured more prominently? Because at the early stage of concerns about climate change, we lacked the means to measure net energy flux, but we did have a long record of surface temperatures going back to the 19th century. Even now, with extensive satellite data, it is not practical to measure energy out with sufficient accuracy to measure the imbalance. First, the imbalance across the planet varies by time, latitude and date. During summer in one hemisphere, that hemisphere has a positive energy balance, while the opposite hemisphere has a negative energy balance. Those two imbalances sum to close to zero, and the difference is the net imbalance. The technology does not yet exist to measure the imbalance to an accuracy of 0.1W/m2, needed to measure the imbalance that the models infer [Hansen et al. 2011].

What evidence do we have that the energy imbalance is real?

Until accurate direct measurements are possible, the best we can do is look for evidence that the effects other than temperature change are happening. Measuring ocean heat content is still a work in progress (with one paper indicating unexpectedly deep increases in temperature [Balmaseda et al. 2013]), and will be a key indicator. The most obvious one that is directly measurable is latent heat of melting. The Arctic has been losing ice rapidly throughout the period when temperature change has supposedly slowed down, indicating that the energy content of the planet is increasing in that region. In the absence of evidence that the energy content of the planet is decreasing anywhere else, that is an indicator that warming is continuing, even as temperatures are not rising sharply.
Arctic sea ice. 2012 set a new record for a low in
summer sea ice for the Arctic, with no special local
conditions or unusually high global temperature.

In any case is it really true that temperature change has paused? Aside from that the Arctic doesn’t care what we think and is losing ice regardless, what do the numbers actually say?

If you take out the influences of volcanoes (particles they spew into the upper atmosphere have a cooling effect), the solar cycle (hint: cycle implies an effect that cancels out in the long term) and the El Niño Southern Oscillation (hint: oscillation implies an effect that cancels out – you guessed it – over the long term), the trend since 2000 is still up and statistically significant [Foster and Rahmstorf 2011].

My references are all from 2011 – except for a 2013 paper on ocean heat content, an area where further work is called for – recent enough to be current, old enough to have been refuted if wrong.

[Balmaseda et al. 2013] Magdalena A. Balmaseda, Kevin E. Trenberth, Erland Källén, Distinctive climate signals in reanalysis of global ocean heat content, Geophysical Research Letters, 40(9) 1754–1759, 16 May (HTML)
[Foster and Rahmstorf 2011] Grant Foster and Stefan Rahmstorf. Global temperature evolution 1979–2010, Environmental Research Letters 6(4) 2011. (online)
[Hansen et al. 2011] J. Hansen, M. Sato, P. Kharecha and K. von Schuckmann. Earth’s energy imbalance and implications, Atmospheric Chemistry and Physics, 11(24) pp 13421–13449, 2011 (PDF)
[Liu et al. 2011] J Liu, JA Curry, CA Clayson and MA Bourassa. (2011). High-resolution satellite surface latent heat fluxes in North Atlantic hurricanes. Monthly Weather Review, 139(9), 2735-2747 (PDF)
[Pierrehumbert 2011] Raymond T. Pierrehumbert. Infrared radiation and planetary temperature, Physics Today 64(1) 2011, pp 33–38. (PDF)

more sciency stuff
A joule (J) is an energy measure. We measure the rate of energy use in watts (W). We also think of watts as a unit of power (intuitively, a more powerful engine burns energy faster). In electricity household consumption measures, a rate of usage is usually in kW, and we measure total energy use in kWh (thousand watts times hours: these are the units you may see on your electricity bill). kWh are just joules scaled to units that you can relate to your own consumption. 1kWh = 3.6MJ. You can derive this conversion from the fact that an hour is 3600 seconds.

Energy to increase temperature of a material is specific heat. I express this here as energy in kJ to raise temperature of 1kg of material by 1°C. We already have this for water. Some numbers, mainly for metals, for comparison:
  • water: 4.2
  • ice: 2.1
  • aluminium: 0.91
  • copper: 0.39
  • gold: 0.13
  • iron: 0.45
  • lithium: 3.57
  • silver: 0.23
From this we see that water has a remarkably high capacity for storing energy. Although common liquids and fluids generally have higher specific heats than metals, water is near the top of the list. This property, combined with the ability of the oceans to mix temperature change faster than conduction allows (soil and rock, for example, can’t move warmth from their surface except by conduction), explains why the oceans, though only about 66% of the earth’s surface, absorb about 90% of any energy imbalance.

Albedo (the fraction of light reflected) varies a lot. Here are some numbers (1 means reflects 100%, 0 reflects nothing):
  • asphalt 0.02–0.12 (fresh asphalt is darker)
  • conifer forest in summer 0.08–0.15
  • bare soil 0.17
  • green grass 0.25
  • sea ice 0.5–0.7
  • new snow 0.8–0.9
Water albedo varies depending on the angle of incident light; a reasonable comparative figure to use versus ice is 0.06. Building long-term ice is aided by snowfalls because snow has such a high albedo, which reduces any tendency to warm from above after snow has fallen.

And in case you were going to blame solar variation for global warming, check out this picture of solar variation measured by satellite (PDF original here). 1998 and 2005 are among the hottest years on record. 1998 was not far off a solar low. 2005 was well on the way down to the the next low. And the current solar cycle, due to peak some time soon if it hasn’t already, is way below the average. We should be experiencing 100-year lows in temperatures if it was only the sun causing climate variability.

Compare solar variation with temperature (you can get the latest version of this from NASA). The satellite data starts in 1978, so ignore the earlier part of the temperature data. 1978-2012 temperatures clearly trend up (and this is even clearer if you do the stats – get the data from NASA and see for yourself if you don’t believe me). The solar cycle on the other hand is just an oscillation and the trend is down not up in the variation. Again, there’s data free for you to download if you want to check.

To make it a bit easier to compare the two, I rescaled the NASA temperature data so it matches the time scale of the incoming solar energy data and chopped it to fit the satellite record. You can do statistical analyses to compare the two but, even without, it’s pretty clear that temperatures have been increasing while the solar cycle has not only doing its usual fluctuations but flattening out. And, remember, over the time that the latest solar cycle has been failing to peak at the same level as previous cycles, we’ve had record loss of Arctic summer sea ice.