Science Briefs

Forcings and Chaos in Global Climate Change

A drought! A big snowstorm! What's causing it? People want something to "blame" for any climate fluctuation that occurs — surely it must be due to "El Nino" or the "greenhouse effect" or "something". A predilection for a deterministic explanation of climate variations is shared by scientists and laypersons.

But climate can fluctuate without help from any such "forcing" mechanism, simply because the atmosphere and ocean are fluids that are always sloshing about. There is no way to predict the timing and location of individual sloshes far ahead of time. This chaotic aspect of climate is a natural consequence of the fundamental "nonlinear" equations that describe the climate system.

Figure 1
Figure 2
These laws of physics permit a range of solutions, and although some solutions are more probable than others, there is no way to know which of the allowable paths Earth will follow. In fact, if we went back in time, like Marty McFly in Back to the Future, Earth's weather and climate would be different the second time we passed through the same period.

In effect, a global climate model lets us fly back to an earlier time and investigate the possible causes of climate variations. A group of scientists, educators and students at GISS, dubbed the "Pinatubo team", has used the GISS climate model to "fly back" to 1979 and investigate the period 1979-1996. The model lets the world run through this period many times, generating an "ensemble" of climate realizations. And it can generate more ensembles in which climate forcings, such as volcanos and greenhouse gases, are added to the model one-by-one to study their effects.

The Pinatubo team found that when the model is subject to measured climate forcings, namely greenhouse gases such as CO2, volcanic aerosols, ozone depletion, and solar radiation changes, the model yields a global average temperature change similar to that observed. This is shown in Fig. 1 for the ensemble of model runs when the ocean temperatures are specified according to observations and the atmosphere fluctuates at will.

Fig. 2 shows the model results when the temperature of both atmosphere and ocean are allowed to fluctuate. The model still does a good job of simulating observed global temperature change. The ocean model employed in this case is the simplest kind, and ocean dynamical fluctuations, such as the observed El Nino warming in 1983 or the La Nina cooling in 1989, are impossible to simulate with this model. Nevertheless, the average of the model simulations is highly correlated with observations (68%).

A principal conclusion of this study is that observed climate change in the past two decades contains unambiguous signatures of both natural and anthropogenic climate forcings. Specifically, observed cooling of Earth's surface in the early 1990s and stratospheric warming in the same period were both a consequence of sulfuric acid aerosols in Earth's stratosphere, a natural climate forcing produced by the Pinatubo volcanic eruption in 1991. In addition, a cooling trend over the period 1979-1996 in the lower stratosphere and upper troposphere, at altitudes 10-20 km, was a result of ozone depletion, which in turn is a consequence of human-made chemicals.

A second conclusion is that year-to-year climate change at middle latitudes, such as the United States and Europe, are primarily chaotic fluctuations. Forcings such as greenhouse gases, on the average, alter temperature over two decades by only a few tenths of a degree, while chaotic fluctuations are several times larger. Thus the forcings, even from a strong El Nino, can modify the probability of unusual temperature or precipitation by a modest amount, but they do not allow for a reliable definitive forecast of seasonal climate. Therefore any claims that El Nino will do this or that to next winter's climate at middle or high latitudes should be taken with a large grain of salt.

Still a third conclusion of this study is an inference that Earth is not in radiative balance with space. Specifically, the observed temperature changes imply that Earth is absorbing about 0.5 Watt per square meter of sunlight more than it is emitting back to space. This imbalance is presumably due to greenhouse gases added to Earth's atmosphere over the past century, to which climate has only partially responded because of the large thermal inertia of the ocean. A consequence of this imbalance is that future warming of about 0.5°C can be expected even if atmospheric composition should remain fixed at today's amounts; i.e., future warming of 0.5°C is already "in the pipeline". It is this slow response of the climate system that complicates the issue of whether and how much greenhouse gas emissions should be restrained.

One inference of this inferred radiative imbalance of Earth is that the ocean is presently accumulating heat, at a rate which should be measurable over several years. Another inference is that, as suggested by Fig. 2, new record global temperatures are likely to occur during the next few years.


Hansen, J., et al. 1997. Forcings and chaos in interannual to decadal climate change. J. Geophys. Res. 102, 25679-25720.