Climate Science Glossary

Term Lookup

Enter a term in the search box to find its definition.


Use the controls in the far right panel to increase or decrease the number of terms automatically displayed (or to completely turn that feature off).

Term Lookup


All IPCC definitions taken from Climate Change 2007: The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Annex I, Glossary, pp. 941-954. Cambridge University Press.

Home Arguments Software Resources Comments The Consensus Project Translations About Support

Bluesky Facebook LinkedIn Mastodon MeWe

Twitter YouTube RSS Posts RSS Comments Email Subscribe

Climate's changed before
It's the sun
It's not bad
There is no consensus
It's cooling
Models are unreliable
Temp record is unreliable
Animals and plants can adapt
It hasn't warmed since 1998
Antarctica is gaining ice
View All Arguments...

New? Register here
Forgot your password?

Latest Posts


The albedo effect and global warming

What the science says...

Select a level... Basic Intermediate

The long term trend from albedo is of cooling. Recent satellite measurements of albedo show little to no trend.  

Climate Myth...

It's albedo

"Earth’s Albedo has risen in the past few years, and by doing reconstructions of the past albedo, it appears that there was a significant reduction in Earth’s albedo leading up to a lull in 1997. The most interesting thing here is that the albedo forcings, in watts/sq meter seem to be fairly large. Larger than that of all manmade greenhouse gases combined." (Anthony Watts)

At a glance

What is albedo? It is an expression of how much sunshine is reflected by a surface. The word stems from the Latin for 'whiteness'. Albedo is expressed on a scale from 0 to 1, zero being a surface that absorbs everything and 1 being a surface that reflects everything. Most everyday surfaces lie somewhere in between.

An easy way to think about albedo is the difference between wearing a white or a black shirt on a cloudless summer's day. The white shirt makes you feel more comfortable, whereas in the black one you'll cook. That difference is because paler surfaces reflect more sunshine whereas darker ones absorb a lot of it, heating you up.

Solar energy reaching the top of our atmosphere hardly varies at all. How that energy interacts with the planet, though, does vary. This is because the reflectivity of surfaces can change.

Arctic sea-ice provides an example of albedo-change. A late spring snowstorm covers the ice with a sparkly carpet of new snow. That pristine snow can reflect up to 90% of inbound sunshine. But during the summer it warms up and the new snow melts away. The remaining sea-ice has a tired, mucky look to it and can only reflect some 50% of incoming sunshine. It absorbs the rest and that absorbed energy helps the sea-ice to melt even more. If it melts totally, you are left with the dark surface of the ocean. That can only reflect around 6% of the incoming sunshine.

That example shows that albedo-change is not a forcing. That's the first big mistake in this myth. Instead it is a very good example of a climate feedback process. It is occurring in response to an external climate forcing - the increased greenhouse effect caused by our carbon emissions. Due to that forcing, the Arctic is warming quickly and snow/ice coverage shows a long-term decrease. Less reflective surfaces become uncovered, leading to more absorption of sunshine and more energy goes into the system. It's a self-reinforcing process.

If you look at satellite images of the planet, you will notice the clouds in weather-systems appear bright. Cloud-tops have a high albedo but it varies depending on the type of cloud. Wispy high clouds do not reflect as much incoming sunshine as do dense low-level cloud-decks.

Since the early 2000s we have been able to measure the amount of energy reflected back to space through sophisticated instruments aboard satellites. Recently published data (2021) indicate planetary albedo, although highly variable, is showing an overall slow decrease. The main cause is thought to be warming of parts of the Pacific Ocean leading to less coverage of those reflective low-level cloud-decks, but it's early days yet.

Albedo is an important cog in the climate gearbox. It appears to be in a long-term slow decline but varies a lot over shorter periods. That 'noise' makes it unscientific to cite shorter observation-periods. Conclusive climatological trend-statements are generally based on at least 30 years of observations, not the last half-decade.

Please use this form to provide feedback about this new "At a glance" section. Read a more technical version below or dig deeper via the tabs above!

Further details

"Clouds are very pesky for climate scientists..."

Karen M. Shell, Associate Professor, College of Earth, Ocean and Atmospheric Sciences, Oregon State University, writing about cloud feedback for RealClimate.

Earth's albedo is the fraction of shortwave solar radiation that the planet reflects back out to space. It is one of three key factors that determine Earth's climate, alongside the evolution of both solar irradiance and the greenhouse effect. Back in the 1990's, the evolution of Earth's albedo was by far the least understood of the three key factors. To address that uncertainty, it was proposed to measure Earth's albedo continuously over at least one full solar cycle. The long data series thereby obtained also helped scientists to explore potential correlations between varying solar activity and albedo change.

Thus was born the Earthshine project. It began in the Big Bear Solar Observatory (BBSO) in California in the mid-1990's. Measuring Earth's albedo was done by making observations of the illumination of the dark side of the Moon at night by light reflected off the dayside Earth. This method was pioneered in 1928 by French astronomer Andre-Louis Danjon (1890-1967).

Trial Earthshine observations were made in 1994–1995 and regular, sustained data-collection commenced in 1998. Data-collection continued until the end of 2017, representing some 1,500 nights spread over two decades.

 Illustration of Earthshine.

Fig. 1: When the Moon appears as a thin crescent in the twilight skies of Earth it is often possible to see that the rest of the disc is also faintly glowing. This phenomenon is called earthshine. It is due to sunlight reflecting off the Earth and illuminating the lunar surface. After reflection from Earth the colours in the light, shown as a rainbow in this picture, are significantly changed. By observing earthshine astronomers can study the properties of light reflected from Earth as if it were an exoplanet and search for signs of life. The reflected light is also strongly polarised and studying the polarisation as well as the intensity at different colours allows for much more sensitive tests for the presence of life. Image and caption credit: ESO/L. Calçada.

In 2005, a new automated telescope was installed in a small, dedicated dome at the BBSO. The two telescopes, new and old, were then run together from September 2006 through to January 2007, for calibration purposes. Observations made with the more accurate automated telescope were then made through to the end of 2017.

Since the early 2000s, scientists have also been measuring planetary albedo with a series of satellite-based sensors known as Clouds and the Earth’s Radiant Energy System, or CERES. These instruments employ scanning radiometers in order to measure both the shortwave solar energy reflected by the planet - albedo in other words – and the longwave thermal energy emitted by it. The overall aim is to monitor Earth's ongoing energy imbalance caused by our copious greenhouse gas emissions.

The Earthshine project and the CERES satellite-based measurements (2001-present day) both record great variation in albedo. That is as might be expected, because cloudiness is such an important albedo-controlling factor and varies so much. However, a slightly decreasing trend was detected (fig. 1, Goode et al. 2021).

Earthshine annual mean albedo anomalies 1998–2017. 

Figure 2: Earthshine annual mean albedo anomalies 1998–2017 expressed as reflected flux in Wm. The error bars are shown as a shaded grey area and the dashed black line shows a linear fit to the Earthshine annual reflected energy flux anomalies. The CERES annual albedo anomalies 2001–2019, also expressed in Wm, are shown in blue. A linear fit to the CERES data (2001–2019) is shown with a blue dashed line. Average error bars for CERES measurements are of the order of 0.2 Wm/2. From Goode et al. 2021.

The data cover two solar maxima, in 2002 and 2014, plus a solar minimum in 2009. Recorded variations in albedo show no correlation with the 11-year solar cycle, the cosmic ray flux or any other solar activity indices. Therefore, the data do not support any argument for detectable effects of solar activity on the Earth's albedo over the past two decades.

In comparison with the CERES data, both show a downturn in recent years, even though they cover slightly different parts of the Earth (Goode et al. 2021 and references therein). To put some numbers on things, in the earthshine data the albedo has decreased by about 0.5 Wm, while for CERES data, 2001–2017, the decrease is about 1.5 Wm. CERES data shows the sharp downturn to have begun in 2015.

The explanation put forward for the difference in albedo decrease between Earthshine and CERES has been further investigated and calibration-drift, a known issue with satellites, has been discounted. Instead, a recent and appreciable increase in sea surface temperatures off the west coasts of North and South America has been cited. The increase has led to reduced overlying low level cloud-deck cover. That would certainly cause significant albedo-decrease. The sea surface warming is attributed to a flip in the Pacific Decadal Oscillation (PDO), beginning in 2014 and peaking during the 2015–2017 period. It began to decline before the end of the decade.

However, a lot of this is very new, as pointed out by Gavin Schmidt at Realclimate in 2022. The role played by, for example, aerosols is not quantified in any great detail yet. But qualitatively, these developments demonstrate how impacts to the long-wave radiation combined with cloud feedbacks can lead to big shifts in short-wave reflectivity. Needless to say, this complex area is the firm focus of much ongoing investigation and will be for the foreseeable future.

Last updated on 3 March 2024 by John Mason. View Archives

Printable Version  |  Offline PDF Version  |  Link to this page

Argument Feedback

Please use this form to let us know about suggested updates to this rebuttal.


Prev  1  2  3  4  5  6  

Comments 126 to 133 out of 133:

  1. blaisct @125,

    Your URLs don't work but...

    I assume when you say @123 that "the psychometric chart in @106 shows the math" you mean that this Free Online Interactive Psychrometric Chart for HVAC engineers does the aritmentic. The "math" is your own. And I would suggest it includes questionable assumptions as well as error and is incomplete. Certainly the contradiction I posed @122 (that Costa et al [2007] showed increasing albedo rather than the required decreasing albedo yet still showed warming temperatures) remains unexplained.

  2. I don't know what went wrong but I tested these; Online psychrometric chart  and Tutorial

    Please except my assumption and online calculator results in the summary of the three conditions, they are only examples.
    Doing some further research on the clouds I found this site Cloud Ceiling Calc that had a correlation used by airplane pilots to predict cloud celling.

    Cloud ceiling (m) = (ground temp. – ground dew point)/2.6 *1000*0.3084
    The base case is virgin land with lots of trees and vegetation, and is simulated by adding water to the online psychrometric chart calculator. The water added (22% of total water) is typical of data for rain forest type land.
    I have added the cloud ceiling calculation to the summary to show that the LHAC cases increase the cloud ceiling no matter what the albedo is. The high albedo case is 20% higher albedo than the base case and much greater than your chart on W/m^2 vs time. The emphasis of the LHAC theory is that cities and cropland do not put as much water into the atmosphere as the orginal virgin land and this lack of water can reduce the cloud cover.

    Summary of these cases:
    Base case: 29.2’C and 70% RH with 23.3 dew point calculating 701 m ceiling
    Low albedo: 33.4’C and 47% RH with 21.4 dew point calculating 1543 m ceiling
    High albedo: 31.5’C and 55.5% RH with 21.4 dew point calculating 1171 m ceiling
    Cloud ceiling and cloud cover should have a correlation.


  3. blaisct @127,

    I'm not sure this interchange is going anywhere. You are not noting the obvious errors in these numbers you are throwing around and if they were corrected I don't see any relevance to the climate change occuring, either globally or regionally.

    On the errors thing, do note that your numbers from the Free Online Interactive Psychrometric Chart are wrong. Consider simplifying the process you are trying to represent. This is not some reversable process so all that matters is the start & end points, not the route between.

    Thus if you choose to start at 25°C & RH=80%, you can add the SH from 16g/kg to 18g/kg (that is 11% increase not 22%) giving RH rising to 89.4% & Enthalpy increasing from 66kJ/kg to 71kJ/kg.
    Now if you add further energy through warming with SH fixed at 18kJ/kg, the enthalpy will rise and the RH will drop with that warming.
    So your Case 3 with an endpoint of 72.3kJ/kg gives a temperature increased from +25°C to +26.2°C & RH drops to 83.6%.
    Your Case 1 with an endpoint of 74kJ/kg gives a temperature increase to +28°C & RH dropping to 75.5%.
    And your Case 2 with reduced albedo giving additional warming to +9.7kJ/kg from the same start conditions yields an endpoint of +29.3°C & RH dropping to 69.7%.

    But these are just numbers. I don't see them relating to what we see of the real world climate change.

  4. Rodger @128
    Once again thanks for your input and patience. My objective in these three cases was to show the difference in air quality (temp and RH) of possible man-made land changes. These air changes are related to the cloud ceiling.
    Sorry for the errors. I do not do a good job going from my excel sheet to this format. I should have shown the before water step in case 1, and I did copy the results of case 1 wrong.
    To correctly compare these cases a base case enthalpy change must be picked based on real world data that represents the middle part of the earth with the sun shining. I have made lots of temp vs RH plots and came up with 8 kJ/kg(da) as a good average change in enthalpy. The same data shows that adding 2g/kg dry air was typical of tropical conditions.
    The short cut you suggested is ok as long as it crosses the 18g/kg water line and the 74kJ/kg(da) (66+8) line simultaneously. The two albedo cases are ether side of the 8kJ/kg base case at 6.4 kJ/kg and 9.7kJ/kg. I corrected the cases to include the case 1 with out water added and the enthalpy difference for each case. All cases start at the same 25’C and 80%RH.
    Cloud ceiling (m) = (ground temp. – ground dew point)/2.6 *1000*0.3084
    I hope all the errors are out of these cases and we can discuss the conclusions.
    1. These simple cases show that the beginning (event 1) of the LHAC theory in @121is valid in that land changes that result in lower available moisture will produce higher temperatures and lower RH air even if the albedo is increased.
    2. This higher temperature lower humidity air is correlated to cloud ceiling.
    Base case water added: typical rain forest (other vegetation or water sources would have less water added)
    Base case no water: Just to show what the rain forest would look like without water added. Note same enthalpy change and same dew point of all the other cases.
    Low albedo: intended to simulate a UHI.
    High albedo: intended to simulate the rain forest conversion in Amazonia.
    Summary of these cases:
    Base case water added: 8kJ/kg(da), 27.9’C, 75.5% RH, 23.3 dew point calculating 561 m ceiling
    Base case no water: 8kJ/kg(da), 32.5’C, 52.0% RH, 21.4’C dew point calculating 1318 m ceiling
    Low albedo: 9,7kJ/kg(da), 33.4’C, 47.0% RH, 21.4’C dew point calculating 1543 m ceiling
    High albedo: 6.7kJ/kg(da), 31.5’C, 55.5% RH, 21.4’C dew point calculating 1171 m ceiling
    Cloud ceiling and cloud cover should have a negative correlation? This exercise also suggests that the LHAC theory is more related to cloud prevention than destruction. The real-world origins of the ceiling correlation to temp and dew point suggest the plume of hot low RH air reaches high into the atmosphere supporting the model in Figure 3 @121.
    Comments on how big (% of earth’s surface) this effect is? See event 2 calculation @121. I get 7.8% of the earth surface that could be affected by hot low RH air to some degree. Figure 2 @121 show a decreasing RH over time, suggesting low RH air is being produce.

  5. blaisct @129,

    The correction of the numbers is good but whether it leads you to anywhere useful is another matter entirely. I repeat my parting comment @128 - "But these are just numbers. I don't see them relating to what we see of the real world climate change."

    Perhaps you should read up on the literature examining the impact of UHI on climate. But be warned, to my understanding there is no evidence suggesting anything but local effects.

  6. I have been watching this discussion for a while, and I too have a really difficult time understanding what blaisct's real purpose or argument is. With respect to albedo, it seems as if he is implying that albedo causes the change in climate, while ignoring the possibility that other factors are changing the climate and albedo is responding to that - the classical albedo feedback that is a standard part of climate science.

    I have access to some high temporal resolution surface radiation data from a continental location. Let's look at four graphs of daily values:

    January radiation and albedo:

    January radiation

    January albedo

    ...and the same location in July

    July radiation

    July albedo

    Let's talk about the last two first. It's a mostly sunny day. with some morning cloud and mid-day scattered cloud. Global radiation peaks at over 1000 W/m2. There is a strong diurnal pattern to albedo - lowest in mid-day (less than 0.2), and highest around sunrise and sunset (around 0.3).

    Then let's compare these to the first two, from January. A similar day in the sense of morning cloud and afternoon clear skies, but global radiation is much lower - (peaks at about 300 W/m2). Albedo is quite different - it drops from about 0.9 in the morning to

    I also know a bit about the temperatures on each day. In July, it was much cooler in the morning and evening, and hottest in the early afternoon. January was much, much colder.

    Should I assume that the differences in albedo have caused those temperature differences? After all, there is a strong correlation: albedo drops, and temperature rises. Very high albedo? Very cold temperatures!

    ...but all I have done is shown that winter is colder than summer, so you can get snow on the ground instead of agricultural crops. After all, the energy input from solar radiation in January peaks at 30% of what it was on that July day, even if we don't account for the higher January albedo and shorter daylight period.

    And the diurnal cycle in July? It is well-known and well-documented that surface albedo shows variability with solar zenith angle in clear skies. The sun is high in the sky at solar noon (which is about 1pm clock time on these graphs), and low in the sky at sunrise and sunset. It's not the albedo that is driving temperature differences: it is the change in solar input.

    Nothing surprising here. Albedo differences are the result of other factors that affect weather and climate.

    I think the same applies to blaisct's humidity and cloud arguments. There is nothing that I can see in his comments that gives any evidence that albedo or humidity are the driving force behind changing climate - they can (and are more likely to be) the result of a changing climate. A feedback, not a forcing.

  7. Bob Loblaw @131,

    I also have struggled to identify any sign of a significant driver of climate in the arguments presented by blaisct. If we wind back to the initial proposal (in the 'Does Urban Heat Island effect exaggerate global warming trends?' thread @59), I feel the scoping of a direct potential forcing can be scoped quite simply** but refining such an analysis does not appear possible with commenter blaisct who now introduces further speculative feedbacks into the discussion, thus piling unhelpfulness on top of unhelpfulness.
    (**According to Wild et at [2015] fig2a, the average land albedo equates to 48Wm^-2(land) = 14Wm^-2(global). If urbaniseation reduced that to zero over 1M sq km, that would equate to a 0.1Wm^-2(global) forcing, thus a maximum value for a quantity which may not even be positive. Note Guo et al [2022] suggest the effect is negative over urbanisation in China.)

  8. Please note: the basic version of this rebuttal has been updated on March 3, 2024 and now includes an "at a glance“ section at the top. To learn more about these updates and how you can help with evaluating their effectiveness, please check out the accompanying blog post @

Prev  1  2  3  4  5  6  

Post a Comment

Political, off-topic or ad hominem comments will be deleted. Comments Policy...

You need to be logged in to post a comment. Login via the left margin or if you're new, register here.

Link to this page

The Consensus Project Website


(free to republish)

© Copyright 2024 John Cook
Home | Translations | About Us | Privacy | Contact Us