Guest post by Willis Eschenbach (@weschenbach on eX-Twitter)
I love science because it surprises me. I have a few today. My first surprise today was evidence of strong negative feedback in surface temperature. Please note that I am not claiming that I am the first to make these observations. I'm just saying, it surprised me.
My approach to scientific research is graphically based. I take large numbers, sometimes tens of thousands, and display them graphically. Sometimes the results are what I expected, even hoped for.
But other times, my latest chart comes on the screen and I say “Wow?” …these surprises make it all worth it. These are the starting points for interesting winding paths. Come take a stroll with me along one of these roads.
Due to a series of misunderstandings and coincidences, I ended up looking at month-to-month changes in the net effect of clouds on radiation. “net effect” Refers to the fact that clouds both warm and cool the surface.
cool down This phenomenon occurs because clouds reflect sunlight back into space or absorb it, blocking it from reaching the ground. Either way the surface will be cooled.
warming Occurs when the portion of thermal radiation emitted by clouds reaches the ground and is absorbed by the ground.
“net effect” It's the difference between two opposing effects — including both, do clouds warm or cool the surface, and by how much?
Unsurprisingly, this is known as the “surface cloud radiative effect”, or “surface cloud CRE”, hereafter referred to as “CRE”. When CRE is negative, it indicates that the net radiative effect of clouds is cooling the surface. Correspondingly, a positive CRE means that clouds warm the surface through radiation changes. Figure 1 shows the 24-year average of net surface CRE recorded by the CERES satellite.
Figure 1. The effect of clouds on the net total amount of radiation (longwave and shortwave) absorbed by the Earth's surface. The horizontal dashed line near the equator marks the edge of the tropics (23.5° N/S). The horizontal dashed lines near the poles are the two polar circles (66.5° N/S). The unit is watts per square meter (W/m2).
Figure 1 has some interesting things.
• Overall, clouds cool the surface by about -19 Watts per square meter (W/m2)
• The ocean cools almost three times more than the land.
• The polar regions of both polar circles are warmed by clouds.
• The only areas warmed by clouds on average are the polar regions and deserts.
• The greatest cooling occurs in the Intertropical Convergence Zone above the equator and the Pacific Warm Pool north of Australia.
However, I have never seen a month-by-month record of surface network CRE. Of course, in order to observe this, we need to observe the two hemispheres separately to avoid the influence of the opposite seasons in the two hemispheres. Figure 2 below shows the monthly changes in the Northern Hemisphere, which was my first surprise.
Figure 2. Northern Hemisphere monthly net surface cloud radiative effect.
I didn't expect the effect to go from a slight warming in the winter to a -40 W/m2 cooling in the summer. This is a huge swing in the cloud effect.
Interestingly, the net cooling effect per decade is -0.2 W/m2. The ten-year increment in CO2 forcing is +0.27 W/m2 (95% CI: 0.22 W/m2 – 0.32 W/m2). Therefore, small changes in surface CRE during the record period are of the same order of magnitude and are opposite (cooling) to any warming effect of CO2 forcing.
Of course, this made me wonder how much different the summer and winter temperatures would be without the radiative effect of clouds… which led me to create Figure 3.
Figure 3. Current Northern Hemisphere summer temperatures (black), and theoretical temperatures without cloud radiation effects (all else being equal, which of course is not the case). Values in all cases are in W/m2 and then converted to temperature using the Stefan-Bolzman equation (assuming an emissivity of 0.95).
Therefore, the average peak Northern Hemisphere summer temperature would not be around 72°F (22°C). Without the differential radiative influence of clouds, the average Northern Hemisphere summer temperature would be around 84°F (29°C). oops! It's also a little colder in winter.
(Yes, I know that without clouds a lot of other things change, so my diagram is purely theoretical. I just want people to understand that cloud cooling from +5 W/m2 in winter is actually from – 40 W/m2.
Out of curiosity, I decided to look at the entire Earth again, as shown in Figure 1, but this time in the Northern Hemisphere in mid-winter (December) and mid-summer (June). Here are these two graphs.
Figure 4. Same as Figure 1, but showing the net surface cloud radiative effect in midsummer and midwinter. December and June averages. Horizontal dashed lines mark the edges of the tropics (23.5° N/S) and the two polar circles (66.5° N/S).
Again, more fun stuff. In mid-winter (December) in the Northern Hemisphere, clouds warm almost all areas north of around 35°N. The same is true in mid-winter (June) in the Southern Hemisphere. Clouds are warm south of 35°S.
Another strange phenomenon. In many cases, white/black contours outline desert areas, where clouds are warming regardless of the season, according to CERES data. Why?
Next, I looked at a scatterplot of surface temperature versus surface cloud radiative effects using 1° latitude x 1° longitude grid cell data. There are 32,400 data points per hemisphere. I graphed the data by season and hemisphere. While doing this, I noticed a strangest phenomenon. This is my second surprise.
The plots of midwinter air temperature and midwinter cloud radiation effects are very similar in both hemispheres.
The same is true for the relationship between Midsummer cloud radiation effects and midsummer temperatures. Summer relationships are similar in both hemispheres. Here are those comparisons.
Figure 5. Grid cell scatter plot. The chart above shows midwinter – midwinter in the Northern Hemisphere (December) and midwinter in the Southern Hemisphere (June). The chart below shows midsummer – midsummer in the Northern Hemisphere (June) and midsummer in the Southern Hemisphere (December).
There are some interesting points here. First, the correspondence between the two winters (above) and the two summers (below) is surprisingly close.
The main difference is the low temperature grid cells in summer. The Southern Hemisphere has open ocean almost all the way to the ice-covered Antarctic Plateau. Whether winter or summer, the clouds in Antarctica are as warm as spring. Therefore, the change in the cloud radiative effect over the Antarctic coastline region during summer is a sudden, almost vertical warming change (left end of orange/black line, lower box). In the Arctic, changes in polar warming are slower and more gradual because the poles are covered by water rather than at higher altitudes in the Antarctic (left end of blue/black line, lower box)
Beyond that, however, the two hemispheres are very similar. Most importantly, cloud cooling rapidly intensifies when temperatures exceed about 26°C in summer and winter, and accelerates as surface warming increases.
For a strange reason, the seasonal similarity of the oceans in both hemispheres is important to me. I used a grid cell-based scatterplot analysis of the type in Figure 5 above, as shown below, to understand the relationship between temperature and CRE on a global scale. See my article Observational and theoretical evidence that cloud feedback reduces global warming for a discussion of the implications of Figure 6 below.
Figure 6. Scatter plot of net surface cloud radiative effect versus surface temperature, for all 1° latitude x 1° longitude surface grid cells.
The main objection people have raised against my use of grid cell-based scatterplot analysis of the type in Figures 5 and 6 above is that they claim it is investigating location based relationship, so it does not prove any direct relationship between the two variables.
Another way of stating the objection is to say that of course there is some given relationship between temperature and CRE for some locations, which relationship is determined by the location-based characteristics of the grid cell in question. Perhaps ocean currents or nearby mountains determine temperature and CRE.
Now, this doesn't seem logical to me, because in Figure 6, the CRE values are grouped by average grid cell surface temperature. There are many grid cells on Earth with very similar average temperatures. But I haven't yet figured out how to rebut this objection to show that it's not location-based.
However, the similarities between hemispheric ocean midwinter and hemispheric ocean midsummer show: The relationship between temperature and cloud radiative effects is not due to some location-specific characteristics.
It cannot be location specific because The two hemispheres have no common location. These are completely different grid cells, in completely different oceans in different hemispheres, with different currents, different depths, different adjacent landmasses…yet the relationship between temperature and surface cloud radiation is surprisingly similar .
So, my third and biggest surprise of the day, after starting down a completely different path, I stumbled across a way to refute my main objection to grid cell based scatterplot analysis .
It's funny how life goes on as I walk down random paths without any guiding star other than an endless curiosity about the wonders of this world.
===
The moon rises over the redwood trees. It must be time for me to go see the moon. I just need a guy in uniform with a Glock to tell me every few hours “Sir, stay away from the computer and keep your hands away from the keyboard so you won't get hurt!”
My best wishes to all of you going forward, my friends—may your lives be filled with wonder, adventure, and surprises of all kinds.
w.
as usual: I politely ask that you quote the exact words you are discussing when commenting. Avoid endless misunderstandings.
Relevant
