Earth’s albedo and its symmetry

In an article appearing in AGU Advances, Dr. George Datseris and Prof. Bjorn Stevens provide an analysis of the Earth’s albedo, its surprising symmetry between the two hemispheres, and how clouds make it possible.

The Earth’s albedo is the portion of the incoming solar irradiance that is reflected back to space by the planet as a whole. Satellite observations provide us with measurements of the solar irradiance balance at the top of the atmosphere and reveal that on average our planet reflects 30% of the irradiance. What is completely surprising is, that, on average, each individual hemisphere of the planet also reflects 30%. This is surprising because land reflects more sunlight than ocean, and the northern hemisphere has more land than ocean when compared with the Southern Hemisphere. So, with this consideration alone, the Northern Hemisphere should reflect more sunlight, i.e., have higher albedo, simply because of this surface imbalance.

The key “actor” for Earth’s albedo symmetry are clouds, who are responsible for about half of the planet’s albedo. It turns out that clouds at the largest scales are distributed in such a way as to compensate the deficit of surface albedo of the Southern Hemisphere. This fact was observationally known for decades now. Unknown is, how clouds do this and why. Yet knowing this answer has implications for fundamental processes underlying the planet’s energy budget as well as cross-equatorial heat transport. 

Datseris and Stevens were able to answer the first part of this question and provide several hints for the second part. To address how clouds compensate for the surface albedo imbalance between the two hemispheres, they constructed a new spatiotemporal field representing cloud albedo using measurements of cloud fraction and optical depth from Clouds and the Earth's Radiant Energy System: (CERES) data, shown below.



Figure 1: Temporally and zonally averaged cloud albedo field C (from Datseris and Stevens, 2021). It is further compared with the cloud fraction f which is typically used to represent cloudiness (normalized to same units). Black numbers show the spatial average of C in each colored latitude band.


The analysis showed that tropical cloudiness is mostly balanced between the two hemispheres, even though the Intertropical Convergence Zone (ITCZ, a large band of cloudiness near the equator spanning most of Pacific and Atlantic oceans) is located mostly north of the equator. In contrast, the cloudiness of extratropical storm tracks, especially over ocean (not shown in the figure) are largely imbalanced, with much higher cloud albedo over the ocean in the Southern Hemisphere storm tracks. This cloud albedo imbalance compensates the imbalance of surface albedo between the two hemispheres, bringing an overall symmetric albedo.

The question that remains now is whether this balancing effect of clouds is by chance, or some “compensation mechanism” underlines the process. To understand what properties this mechanism should satisfy if it exists, Datseris and Stevens analyzed reflected solar irradiance data from CERES directly. Specifically, they analyzed the de-seasonalized, hemispherically averaged timeseries R' (also called anomalies), shown in the following figure.


Figure 2: De-seasonalized hemispherically averaged timeseries for reflected shortwave radiation at the top of the atmosphere, for north (teal) and south (black, dashed) hemisphere (from Datseris and Stevens, 2021).


Even though the timeseries themselves are indistinguishable from noise, they are dominated by an identical long-term trend spanning decades. Whereas as the northern hemispheric trend has been attributed to changes in stratocumulus clouds over the north-east Pacific, what came as a surprise was a similar southern hemispheric trend. This hints at a physical mechanism, and if one exists the properties of the time-series help narrow the range of time-scales on which it might operate. The lack of correlation on monthly timescales (the data in this regime are noise with no dynamic indication), rules out short-timescales. The compensation on the time-scales of the time-series rules out long-timescales  (as if a mechanism operated on multi-decadal scales, then there would be no reason that the trends of the two hemispheres match so perfectly). The remaining candidate timescales for a compensation mechanism are therefore between a year and a decade.

The publication was selected for featuring as an Editor’s Highlight.

Original publication:

Datseris, G. and B. Stevens (2021) Earth’s albedo and its symmetry. AGU Advances.



Dr. George Datseris
Max Planck Institute for Meteorology
Phone: +49 (0) 40 41173 308
Email: george.datseris@we dont want