Where The Temperature Rules The Total Surface Absorption

By Willis Eschenbach – Re-Blogged From http://www.WattsUpWithThat.com

Reflecting upon my previous post, Where The Temperature Rules The Sun, I realized that while it was valid, it was just about temperature controlling downwelling solar energy via cloud variations. However, it didn’t cover total energy input to the surface. The total energy absorbed by the surface is the sum of the net solar energy (surface downwelling solar minus surface reflections) plus the downwelling longwave infrared, or DWIR. This is the total energy that is absorbed by and actually heats the surface.

According to the CERES satellite data, globally, the solar energy absorbed by the surface averages 162 W/m2. The downwelling longwave averages 345 W/m2. Conveniently, this means that on average the earth’s surface absorbs about a half a kilowatt per square meter on an ongoing basis. (And no, I have no interest in debating whether downwelling longwave radiation actually exists. It’s been measured by scientists around the world for decades, so get over it, Sky Dragons. Debate it somewhere else, please, this is not the thread for that.)

Let me note in passing that a doubling of CO2, which will increase the DWIR by something on the order of 3.7 W/m2, and which it is claimed would lead to Thermageddon, would be less than a 1% change in total downwelling radiation at the surface … which would easily be offset by a small change in total cloud cover. But I digress.

Here is the correlation between temperature and total surface absorption.

ceres total surface absorption vs temperature.png

Figure 1. Correlation of total surface absorption with temperature.

Note the similarity to the previous graph showing just the correlation between surface temperature and downwelling solar energy at the surface.

Now, to explain how this can happen I need to take another digression. I was attracted to the study of the climate, not by questions about why the temperature was changing so much, but by why it was changing so little. As a man with some experience of heat engines and governors, I found it amazing that the temperature of such a possibly unstable system could only have changed by ± 0.3°C over the entire 20th century. Why should such a world, with clouds appearing and disappearing, with huge volcanoes popping off every few decades, with winds going up and down, with storms and hurricanes appearing and vanishing, why would it be so stable in the long-term? So I started looking for some long-term kind of feedbacks that could explain it.

I was living in Fiji at the time. After literally months of fruitless searching and thinking about long-term slow feedbacks, one day I thought “Hang on. I’m looking at the wrong end of the time spectrum.” What I realized was that if there was something that kept the daily temperatures from going outside a certain range, that would, in turn, keep the weekly, monthly, annual, decadal, centennial, and millennial temperatures from going outside that same range.

And because I was living in Fiji, the answer was right above me. The daily tropical weather typically looks like this: clear at dawn, clouding up with thermally-driven cumulus clouds in the late morning, perhaps thunderstorms in the afternoon if the day is warm enough, clearing some time after dark. Lather, rinse, repeat, as they say.

I also realized that there were two variables in that scheme—the time of onset of the cumulus clouds and the thunderstorms, and the amount of each of them. I hypothesized that these factors were what controlled the tropical temperature. Since then I have amassed a lot of evidence that my hypothesis is correct, including this post and its predecessor.

There are some important things to note about this process. First, the time of the emergence of the cumulus fields and the thunderstorms is NOT dependent on total forcing. Instead, they are responding to surface temperature. When the surface is cool at dawn, clouds form later, and more sunlight comes in, warming the surface. When the surface is warmer, clouds form earlier, throttling the energy input to the system, and cooling the system back down.

As a result, the system is not affected by small changes in insolation. For example, if a volcanic eruption reduces the amount of sunshine making it through the stratosphere, the tropics cool. And when they cool, clouds form later, letting in more sunlight, and rebalancing the system.

Next, the response is based, not on average temperatures, but instantaneous temperature. As such, it is obscured by monthly or yearly temperature averages.

Finally, the response is immediate. There is no lag of days, weeks, or months. As soon as the temperature crosses some given threshold, clouds form immediately, cooling the surface. This effect is so powerful that although the morning sun is growing stronger and stronger, when the clouds kick in, the temperature can actually drop. Here’s a graph of the long-term average daily swings of a number of TAO buoys spread across the Pacific. Here are the locations of the buoys. I’ve used those on the equator because they have the most data. The TAO buoy data is available hereTAO Buoy Locations

Figure 2. Locations of the TAO buoys

These readings were taken by the automated buoys every ten minutes.

TAO daily cycles temperature

Figure 3. Daily average temperatures, equatorial TAO buoys.

In the cooler areas at the bottom of the graph, the onset of the morning cumulus field merely slows the daily warming. But in the warmer areas, when the clouds appear, the temperature actually drops. The differences can be seen clearly when they are expressed as anomalies about their individual average values, viz:

TAO daily cycles temperature anomalies

Figure 4. Daily temperature anomaly variations, equatorial TAO buoys.

Note that this “overshoot”, the ability to drive the temperature below the local initiation temperature threshold, is critical to controlling a lagged system such as the climate. It is also present in thunderstorms. They generate their own fuel once they are started, allowing them to cool the surface below the initiation temperature threshold.

Next, I divided the days into those which were warmer than usual from midnight to 5 AM, and compared them with the days which were cooler than usual during that same time span. Here’s the result:

TAO average warm and cool days warmest buoy

Figure 5. Averages of warm and cool days, one of the warmest TAO buoys

This shows the temperature control in action at one of the warmest TAO buoys. On days which start out warmer than normal, the clouds and thunderstorms form earlier and more strongly. By evening the temperatures cool towards the average value. The opposite happens when the temperature from midnight to 5 AM are cooler than usual—cumulus form later and more scattered, thunderstorms may not form at all. And as a result, the surface warms towards normal.

With that understanding, we can take another look at the graphic in Figure 1, which I reproduce here:

ceres total surface absorption vs temperature.png

Consider that this is a long-term average. This means, for example, that temperatures in the green and light yellow areas immediately outside the gray lines are not really slightly correlated with the total downwelling radiation.

Instead, it means that the number of days during which they are negatively correlated is slightly less than the number of days when they are positively correlated. However, this average conceals an important fact—the negative and positive correlations are not randomly distributed.

Instead, emergent phenomena like cumulus fields and thunderstorms occur earlier and more strongly exactly when and where the surface is hot. So those areas around the gray outlines of negative correlation are doing the same thing as the areas within the gray outlines—cooling down the hottest days and warming up the coolest days. The only difference is that the warm days are less frequent than inside the gray outlines. This puts limits on how much analysis we can do using averages, as I highlighted in “The Details Are In The Devil“.

In conclusion, let me say that the emergence of the tropical cumulus fields and associated thunderstorms are not the only temperature-linked phenomena which participate in global temperature regulation. Other phenomena include dust devils, squall lines, the Atlantic Multidecadal Oscillation, the El Nino-La Nina pump, cyclones, and the Pacific Decadal Oscillation. Likely more as well …

Me, I’m sitting on a hill in the Solomon Islands on what is scheduled to be my last day here … you’re welcome to read about it, along with the story of the Crocodile and Tufala Panadol over at my blog, Skating Under The Ice.

Best of life to all,




Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s