Earth’s Energy Imbalance: Quarter century update
Data from the NASA CERES satellite system is now available to the end of 2025. What can it tell us about warming rates, trajectories and the engagement of feedbacks?
The Earth’s Energy Balancing Act
Most discussions about global warming and climate change hinge on temperature increases, be it surface air temperature, sea surface temperature or ocean heat content. There is another common way of assessing what is going on though, and that is to analyse the observational energy flow data measured by satellites. In particular the NASA Clouds and the Earth’s Radiant Energy System (CERES) mission.
Since March 2000, the project has been measuring the energy flows of the planet and its cloud properties. The data is freely available online for everyone to check out.1
At the end of the day, whether the Earth warms, stays stable or cools is down to the flow of energy from and back into space. Solar radiation comes in, typically at about 340 Watts per square meter (W/m2) of the Earth’s surface. Some of this is reflected back out into space by clouds and reflective surfaces like snow and ice, although every surface has some reflectivity known as its level of Albedo. Typically about 29% of the incoming solar radiation is reflected in this way, leaving the rest to be absorbed, providing the ‘Energy In’ side of the balance.
If that was all, the Earth would heat continually and would now be a red-hot crisp, which clearly is not the case, so energy must also escape somehow. It does this through infra-red radiation back out into space. All objects radiate long wave infra-red energy depending on their temperature. That’s how night-vision cameras and PIR detectors work, their electronic sensors are sensitive to the infra-red light that we can’t see. The higher the temperature of the surface, the more energy is emitted. Some snakes can “see” in infra-red allowing them to hunt rodents in the dark.
As the Earth is warmed, it responds by emitting infra-red energy. If the amount of this long wave radiation leaving the Earth is equal to the absorbed solar radiation coming in, the system is in balance and a steady state exists.
Greenhouse gasses (GHG) come into play by preventing some of the long wave radiation from escaping. If there were no greenhouse gases at all, the average temperature on Earth would be about 30ºC lower, so they do a good job at making the planet hospitable. There are a range of GHGs but the most important is carbon dioxide. It’s important because it has such a long lifetime in the atmosphere which means its level acts as a planetary temperature control knob.
Balance is achievable at any GHG level. Say we increase CO2 levels from the pre-industrial level of 278ppm to 556ppm, a doubling, then leave it at that level to let the Earth find equilibrium. Temperatures would rise because the energy imbalance would be positive, energy in would be more than energy out as the GHG would trap more outgoing long wave radiation. As the temperature rose though, more infra-red would be emitted from the warmer surfaces until a balance was found where the amount of energy being radiated was sufficient to overcome the increased GHG retention and exit the atmosphere in balance with the energy input.
The value of this temperature rise for a doubling of CO2 is called the Equilibrium Climate Sensitivity (ECS) and is one of the more important, but also most difficult to pin down, values in climate science. The IPCC has a range of between 2º and 5ºC but recent observations suggest a value of 4ºC or more. So over time, the temperature would rise and stabilise at +4ºC at the end of our doubling experiment. The value of ECS is hard to pin down as there are so many different feedbacks and responses within the global climate system that occur at different temperatures, such as the amount of ice, cloud properties, weather effects, vegetation and nutrient cycle changes etc. that all play against each other until an equilibrium is reached.
Today’s Earth Energy Imbalance (EEI)
The CERES mission provides all the data we need to examine this energy balancing act and derive an important value - that of the Earth’s Energy Imbalance (EEI). If the value is positive then the Earth is absorbing energy and therefore warming. If it is negative, the Earth is losing energy to space and therefore cooling.
Here is the latest data up to the end of 2025. The EEI is currently 1.44 W/m2. That doesn’t sound like a lot until you remember that the Earth has a lot of square meters. This amount of energy imbalance is actually huge. In terms of heat energy it is the equivalent of over 11 Hiroshima bombs worth of energy being added to the climate system every second! That is based on a 36 month running mean to smooth out the noise, but value for 2025 alone was even higher at 1.48 W/m2.
Not only is the EEI positive, but more worryingly, it is steadily increasing. For the first 36 months of the century it was just 0.34W/m2, so it has increased by over four times in just 25 years! Since EEI is effectively the rate of energy accumulation, this results in a clear acceleration of the energy being absorbed by the climate systems. This can only manifest itself as the acceleration in global warming that is now being observed. The Earth Energy Accumulation curve below shows this clearly. In the first 25 years of this century the Earth has accumulated over 345 Zettajoules (1021J) of heat energy. The best way to describe how big a zettajoule is, is to paraphrase Douglas Adams, “it’s big, very big!”
The result of this accelerating energy accumulation is evident in all the observational measures we have for global warming and climate responses. Surface air temperature, sea surface temperature, ocean heat content at every depth, ice sheet melt, glacial retreat, extreme weather frequency and impact, even mosquito born viral infections, are all accelerating in response to the increasing energy flow into the climate system.
Where does all the energy go?
Most climate news focusses on surface air temperature, quite rightly since that is where we live and where the weather events happen, but only 2% of the added energy goes to warming the atmospheric air. The vast majority, 89% goes to warming the oceans (52% warms the top layers to 700m deep, 30% the intermediate layers from 700m to 2,000m and 8% to the deep ocean below 2,000m). 4% goes to melting the ice of the world including the major ice sheets, glaciers and permafrost. That leaves 5% which warms the land surfaces.2
Earth’s Energy Imbalance ladder
We can break down the various components so see where heat is being absorbed and rejected using a ladder chart that shows the changes in EEI elements since 2000. The chart starts on the left hand side with the EEI in 2000 at 0.34 W/m2 and ends on the right with the 2025 level. The short and long wave contributions to the 25 year change are separated into clear and cloudy sky response, resulting in todays EEI of 1.44 W/m2.
The method is based on a 2021 paper ‘Radiative Energy Flux Variation from 2001–2020’ published by Dübal and Vahrenholt who also used the CERES data.
Let’s start with the input to the climate system, Incoming Solar Radiation. This varies on an 11 year solar cycle but due to phasing over the last 25 years has had a slight positive overall contribution with the latest solar maximum accounting for an extra 0.13 W/m2 by December 2025 compared to 2000. That’s 0.04% of the 25 year average of 340.2 W/m2, so no, the sun’s radiation cycles are not responsible for climate change.
Short wave clear sky reflection is reducing thus adding to the EEI. The trend is increasing as cloud cover reduces and the amount of snow and ice also reduces. Short wave clear sky reflection is also reducing due to lower air pollution. Aerosols in the air from fossil fuel burning and industrial outputs reflect sunlight, so as countries improve air quality, reflected radiation reduces and absorbed radiation increases.
Short wave cloudy sky reflection is reducing even faster adding more to the EEI. The trend is increasing as the cloud fraction reduces, cloud structure changes and aerosol pollution reduction diminish their brightness. Cloud area fraction has dropped by 0.31% in the 25 years of CERES measurements to just over 67%. Again that doesn’t sound much, but it equates to an average reduction of 1.58 million km2 of cloud cover! That’s roughly the surface area of Mongolia.
Long wave clear sky radiation should be providing cooling and up until 2020 it was doing so. It has now flipped to providing some additional heating. The trend is for this to continue as more and more GHGs accumulate in the atmosphere. It is likely this aspect will temporarily return to zero or even be positive as the last year has seen a strong increase in output, probably due to the surface heating experienced in the last three record warm years, which have averaged more than +1.5ºC, being slowly radiated upwards.
Long wave cloudy sky radiation is the only remaining cooling factor and fortunately is also strengthening as clouds warm. This is not sufficient to compensate for the other factors however, leading to the continued increase in EEI and continued acceleration of EEA.
Albedo decline
Since the majority of the EEI growth is due to shortwave reflection loss or lowering Albedo, it’s worth looking at this trend in more detail.
The drop is 0.53% in the last 25 years (1.8% of the 2000 level). Again that doesn’t sound like a big change but in terms of energy input, it equates to 1.77 W/m2 of additional energy which is huge. In fact it’s the same energy as the heat from 14 Hiroshima bomb blasts per second.
The fact that the loss of albedo seems to be accelerating is of particular concern since its obviously a key driver in the overall acceleration of the EEA, global warming and the higher level of Equilibrium Climate Sensitivity. A small proportion is due to reduction in snow and ice, particularly sea ice, but the majority is due to cloud changes, both cover and brightness. The concern is that these are feedbacks which will continue to decline as temperatures continue to rise. If this is the case, then there is a lot more warming ‘in the pipeline’ before an equilibrium is reached for the current CO2 concentrations. It’s observations like this that make some scientists believe that +3ºC by 2050 is entirely possible.
Is there anything we can do?
Looking at this observational data, especially the breakdown in the ladder diagram, the two changes that are required to reduce the EEI and stabilise the climate are to increase the amount of long wave radiation that can escape through clear skies, and increase the amount of shortwave radiation that is reflected back out into space.
First and foremost, our emissions of GHGs have to slow and stop as soon as possible. Every tonne we add makes the problem worse and the solutions more difficult.
The following two actions form the pillars of likely future climate policy. Carbon Dioxide Removal (CDR)3 and Solar Radiation Modification (SRM)4. The first is common across most country’s climate and net-zero plans, consisting of natural rejuvenation of forests and mechanical carbon capture and storage. The second, SRM, is more controversial.
Under CDR, CO2 is removed from the atmosphere and stored permanently, preferably in the lithosphere, where it must remain out of the atmosphere effectively forever. It also includes reducing methane emissions which, due to the higher greenhouse gas potential, but short lifetime, are responsible for up to 30% of the current warming. The scale of the challenge is huge. If a potential process or technology can’t be scaled to capture at least a billion tones of carbon a year, it’s not really worth talking about.
Solar Radiation Modification or Geoengineering is starting to be talked about in the open. There are a range of techniques which all try to reflect sunlight back into space to replace the lost cloud and ice albedo. It’s not a solution as it doesn’t remove carbon from the atmosphere, so does nothing to help with ocean acidification for example. If it’s stopped, the EEI would rocket back up to where it would have been, driving a very rapid temperature compensation called a ‘Termination Shock’.
If Climate Change were a cancer, CDR is the chemotherapy - if caught early, the treatment may cure the disease (permanently lower temperatures), it may avoid spreading (avoid triggering tipping points) and death (unsustainable economic and societal damage caused by self-reinforcing feedbacks and a hothouse trajectory).
SRM is the drugs used to reduce the side effects and to make life tolerable and as pain free as possible while the treatment (bringing CO2 down to safe levels) is underway. On it’s own it’s as good as aspirin, it would provide temporary relief but not stop the disease, which would continue to spread.
Both CDR, and SRM in combination, will become more politically likely if the warming damages continue to accelerate, as the EEI data strongly suggests it will. CDR is essential to solve the problem we have created, SRM could help avoid tipping points while it is carried out. SRM without net-zero and CDR will be a waste of time, merely kicking the can down the road. It will be interesting to see what our glorious, courageous, forward-looking politicians opt for when they finally get their heads out of the sand.
NASA CERES Site: https://ceres.larc.nasa.gov/
von Schuckmann, K., et al. Heat stored in the Earth system 1960–2020: where does the energy go?, Earth Syst. Sci. Data, 15, 1675–1709, https://doi.org/10.5194/essd-15-1675-2023, 2023.
The Case for Carbon Dioxide Removal Technologies
The only way to limit temperatures to +1.5ºC, or even +2ºC, by the end of the century is to remove and permanently store a significant quantity of the greenhouse gases accumulating in the atmosphere. To manage an overshoot, where temperatures exceed these limits and are brought back down by 2100, will require even more carbon dioxide removal.
We need to start talking about Solar Radiation Modification
Fifteen years ago the Royal Society published a report on Geoengineering the Climate. The UK Government is now about to fund a series of climate cooling research projects, adding to the growing interest from Universities, philanthropists and venture capitalists around the world. This article looks at what’s involved, why it is needed, how it could be de…
Harris's analysis is largely right. He correctly identifies that the surge in Earth's Energy Imbalance (0.34 → 1.44 W/m² in 25 years) is driven overwhelmingly by albedo decline rather than by extra greenhouse trapping. He spells out that shortwave reflection loss is the dominant ladder rung, that cloud area has shrunk by roughly 1.58 million km², and that "the majority is due to cloud changes, both cover and brightness." His ladder diagram, his framing of W/m² as the right yardstick, and his alarm about acceleration are all on solid ground.
The hiatus is what he leaves out: the cooling capacity of the biosphere itself.
Having diagnosed clouds as the problem, Harris never asks why the clouds are dimming. He gestures at two causes — less sea ice and less industrial aerosol pollution — then jumps to solutions: Carbon Dioxide Removal (slow) and Solar Radiation Modification (fast but risky). The entire biological side of cloud formation is simply absent. Yet clouds need condensation nuclei, and a large share of those nuclei are biogenic: forest-emitted volatile organic compounds over land, and dimethyl sulfide (DMS) from phytoplankton over the oceans. Marine DMS production has fallen sharply since industrialization, and that decline is itself a forcing on the same order of magnitude as deforestation CO₂. A meaningful chunk of the very albedo collapse Harris is alarmed about is a biological signal, not just a cryosphere or industrial-aerosol signal.
This matters most in the tropics, and Harris's global-mean framing obscures that. The tropical zone receives the lion's share of incoming solar radiation — sun nearly overhead year-round, minimal seasonal dimming — so it is the place where small changes in surface energy partitioning translate into large changes in global heating. Tropical rainforests, mangroves, savannas, and the phytoplankton-rich tropical and subtropical oceans evolved precisely as machinery for handling that flux: dense canopies that pump prodigious volumes of water aloft, daily convective cloud towers that reflect sunlight before it ever reaches the ground, and biogenic aerosol production that keeps marine stratocumulus decks bright. Destroying these biomes doesn't just remove a carbon sink — it dismantles the planet's primary heat-dissipation system at its highest-energy latitudes. Every hectare of Amazon converted to pasture, every coral-reef-and-mangrove coastline degraded, every stretch of tropical ocean depleted of phytoplankton heats the planet faster than the equivalent loss at higher latitudes, because the energy throughput there is so much greater.
This is where the Bowen ratio becomes essential. It is the ratio of sensible heat flux (heat that directly warms the air above a surface) to latent heat flux (heat carried aloft as water vapor through evapotranspiration, only releasing when the vapor condenses higher up). It is the single number that determines whether incoming solar energy stays at the surface to bake the boundary layer or gets pumped upward, often above the bulk of the greenhouse layer, where it can radiate to space more efficiently.
The contrast is dramatic. A forest canopy receiving the same solar input as bare ground routes roughly 100 W/m² into evaporation and only ~15 W/m² into sensible heat — a Bowen ratio around 0.15. Bare or degraded ground reverses that: ~15 W/m² latent, ~60 W/m² sensible — a Bowen ratio of 4 or more. Same planet, same energy budget, completely different fate for the energy. Forest vs. bare ground differs by 2–3 W/m² in surface energy partitioning, comparable in magnitude to the entire accumulated fossil fuel forcing signal — and in the tropics, where insolation is highest, the absolute effect is larger still.
Stacked across the five forcing layers — fossil CO₂, deforestation CO₂, soil carbon loss, ocean degradation, and biological cooling loss — our calculations attribute roughly 45% of total anthropogenic heating to the degrading biosphere. That is not a rounding error on the fossil-fuel story; it is nearly half the problem, and crucially it is the half that is restorable on years-to-decades timescales rather than the centuries CO₂ requires.
Harris uses W/m² as his common currency but applies it only to top-of-atmosphere radiative balance. Extended consistently to the surface — and to the Bowen ratio that governs how that surface energy is divided — it surfaces a third lever sitting between his slow-CDR and risky-SRM options: restoring the living systems, especially in the tropics, that brighten clouds, pump latent heat aloft, and seed precipitation. He diagnoses an albedo emergency correctly; he just doesn't see that we dismantled a large piece of the biological cooling machinery ourselves, and could rebuild it.
Climate is a complicated subject with so many interconnected systems.
Great post Tom.
At the end of the day we need some strong actions.