My Oh Miocene

Guest Post by Willis Eschenbach

It was hot here a couple of days ago. I walked past a huge aloe vera plant, taller than my head, that grows by our house. The heat radiating off of the plant was palpable. I could feel a wash of warm air over me as I stood downwind of it. For a while I thought about the curious ability of plants to heat the air around them, and then I let it go.

Figure 1. Solar absorber, natural style. Note how the design efficiently intercepts sunlight because of the spiral, uneven pattern of the leaves. Also note that the design keeps photons from escaping through the funnel-shaped nature of the leaf pattern. Finally, consider that when the plant emits IR from the inner leaves, it will be re-absorbed by outer leaves, perhaps a number of times. This gives the plant, in effect, a local “greenhouse effect” due to the multiple re-absorption of the IR. The leaf geometry also greatly slows down the passage of the air through the plant, once again increasing the local warming. The net of all of those is a warm plant, surrounded by warm air.

I was reminded of our aloe vera again when a friend sent me a copy of the paper “A Warm Miocene Climate at Low Atmospheric CO2 levels,” by Knorr et al. It reports the results of a climate model analysis of the Miocene, the period from about twenty-three million years ago up to five million years ago. It is in press at GRL (paywalled), but the results are discussed here.

In their abstract, we find (emphasis mine):

In this study we present climate simulations of the Late Miocene (11-7 Ma) with a preindustrial CO2 level, using a coupled atmosphere-ocean general circulation model (AOGCM). The simulated global mean surface temperature of ~17.8 ºC represents a significantly warmer climate than today. We have analyzed the relative importance of tectonic [shape and location of the continents] and  vegetation changes as forcing factors. We find that the strongest temperature increase is due to the Late Miocene vegetation distribution, which is more than three times stronger than the impact induced by tectonic alterations. Furthermore, a combination of both forcing factors results in a global temperature increase which is lower than the sum of the individual forcing effects. Energy balance estimates suggest that a reduction in the planetary albedo and a positive water vapor feedback in a warmer atmosphere are the dominating mechanisms to explain the temperature increase. Each of these factors contributes about one half to the global temperature rise of ~3 K. Our results suggest that a much warmer climate during the Late Miocene can be reconciled with CO2 concentrations similar to pre-industrial values.

In looking at the effect of plants on the climate, I’d like to discuss the use of the models, how much weight we should put on their results, and how they could be improved.

The first rule of models says

All Models Are Wrong, But Some Models Are Useful

Their usefulness, of course, depends on their ability to replicate the reality which they are modeling. One issue with the models is that many of them still are not what I call “lifelike”. I discussed this problem of “lifelike” climate model results here. If the models do not act like the real climate, why should we believe them? Unfortunately, no one has ever instituted this kind of test to compare all of the models. It should be a part of a standard suite of climate model tests … dream on.

So at the moment we don’t know if the climate model used in this test gives a lifelike simulation of today, much less of ten million years ago. But I digress. The study says (emphasis mine):

We utilize the comprehensive AOGCM ECHAM5-MPIOM without any flux corrections [e.g. Jungclaus et al., 2006]. The atmosphere model ECHAM5 was used at T31 resolution (~3.75º) with 19 vertical levels. The ocean model MPIOM was run at an average resolution of ~3º with 40 vertical layers. Vegetation is a fixed factor represented by specifying different land surface parameters like albedo, roughness length, vegetation ratio, leaf area index and maximum soil water capacity.

Here we run into another modeling problem. They have set up the vegetation parameters to coincide with what we know of the Miocene landscape. This, of course, means that they are using vegetation as a forcing, rather than a feedback.

But we have been informed, over and over, that the vegetation is a feedback and never a forcing …

This is both a strength and a weakness of the models. We can make assumptions like where the vegetation grew and force things in the model to be a certain way. Then we can see what the effect of that on the results might be.

Unfortunately, the climate doesn’t work that way, where one thing holds steady while everything else changes. So even though we can get some insights, we have no assurance that the effect that we find is real. For example, we don’t know if the Miocene vegetation (which is specified) fits with what the model says were the climate patterns of that time.

Setting aside the manifold questions about the model, there were a couple of interesting parts of the study. The first was that they find that the main effect of the plants occurred through a change in the albedo, particularly for the Sahara. This is in accord with my experience of the aloe vera plant, where it was absorbing much more energy than the ground around it. In part this was because of the albedo of the plant being lower than the ground beneath, but in part it was from the geometry of the plant. (This latter effect is neglected in the model.)

The second interesting thing involves these two statements of theirs about the albedo:

The planetary albedo in MIO [the Miocene simulation] is reduced by ~0.014, which causes less shortwave reflection by the atmosphere and a warming.

and

Based on a zero-dimensional energy balance model [e.g. Budyko, 1969] the impact of α [albedo] and ε [effective long wave emissivity] can be quantified, each causing about one half of the global warming of ~3 K.

Assuming the same solar intensity as the present (345 W/m2), which the authors say that they have done, this change in albedo would result in a change in solar radiation of 0.014 times 345 = 4.83 W/m2. Given the temperature change of 1.5°C from the albedo change, this gives a climate sensitivity of:

1.5°C * 3.7 W m-2 per doubling_CO2 / 4.83 W m-2 = 1.15°C per doubling of CO2.

Me, I think that climate sensitivity is an illusion based on a misunderstanding of how climate works … but for those who believe in it, using Knorr et al’s figures and their concepts, that gives a very low sensitivity, well below the IPCC canonical figure. The IPCC AR4 Summary for Policymakers says (emphasis mine):

The equilibrium climate sensitivity is a measure of the climate system response to sustained radiative forcing. It is not a projection but is defined as the global average surface warming following a doubling of carbon dioxide concentrations. It is likely to be in the range 2 to 4.5°C with a best estimate of about 3°C, and is very unlikely to be less than 1.5°C.

“Very Unlikely”, in IPCC jargon, means less than 10% chance that the sensitivity is less than their minimum estimate of 1.5°C per doubling of CO2. Despite that, this study shows a sensitivity of about three-quarters of the IPCC minimum estimate …

So you’d think that the media headline from this study would be

“Climate Model Finds Extremely Low Climate Sensitivity”

Sadly, that might happen, but only in an alternate universe …

Best to all,

w.