Broadcast meteorologists do not have the best of reputations for predictive accuracy. Audiences are particularly good at remembering – and at pointing the finger – when they’re wrong. Few heap praise when their forecasts turn out to have been accurate.
So the rainy day expected tomorrow sometimes turns out to be sunny, and projections more than a week away are usually offered – and taken – with the proverbial grain of salt.
Given the chaotic elements in weather systems that defy simple calculated predictions, the public understandably asks, How can we forecast Earth’s climate a century from now if we can’t even predict tomorrow’s weather?
The answers lie in important distinctions between weather and climate.
Weather is chaotic (see realclimate.org article on this topic). Many are familiar with chaos theory, often caricatured by the metaphor that a butterfly flapping its wings in Brazil can lead to a tornado in Texas.
Chaos theory states that the outcomes of non-linear dynamic systems are highly dependent on the initial conditions. American mathematician Edward Lorenz, who popularized chaos theory, discovered that rounding-off seemingly insignificant digits in computer weather models resulted in dramatically different weather outcomes. That highly sensitive dependence on initial conditions tends to make weather predictions increasingly difficult as they get further away in time. As a result, reasonably accurate weather forecasts currently are limited to about one week.
Nonetheless, despite being chaotic, weather generally stays within certain defined bounds. No matter what the initial state, it is very unlikely that we will end up with a hurricane off the coast of California. Similarly, we can expect certain places to be rainier at some times of the year than at others. For example, it is impossible to perfectly predict when the monsoon will arrive in India, but scientists can confidently predict that it more likely will arrive in June than in January. The very fact that average winter temperatures in temperate Northern Hemisphere regions can be unambiguously expected to be colder than summer temperatures suggests that a high degree of chaos does not characterize Earth’s long-term climate.
This distinction is important: Scientists cannot predict whether it will be raining or sunny in Stockholm on July 16th, 2075, but they can predict that it will almost certainly be warmer than during the prior December. Similarly, they can predict that the year 2075 will likely be warmer worldwide than the year 2008, and that the decade from 2070-2080 will very likely be warmer than the decade from 1998 to 2008. The longer the timeframe and the lower the spatial resolution, the less likely the predictions will be affected by chaotic behavior.
In the long run, basic components of radiative forcing – atmospheric greenhouse gas concentrations, aerosols, total solar irradiance, and Milankovitch cycles – are the primary determinants of Earth’s climate. The science community’s predictive ability is limited only by the ability to model the magnitude of these forcings and their feedbacks.
General circulation climate models reflect these distinctions. Each model run varies widely in predicted cloud patterns or temperatures for specific days or weeks based on variances in the initial conditions of models. But the models tend to agree on long-term trends and on annual or even monthly temperature anomalies. They cannot predict exactly when it will rain (see NASA GISS Science Brief), but they can predict that average annual precipitation may increase in some regions and decrease in others.
These distinctions between meteorology and climatology underscore the considerable danger in conflating short-term chaotic noise with long-term trends. While climate change is predicted to increase the intensity of hurricanes, for example, scientists cannot say with certainty that any single extreme storm event is the result of climate change. Similarly, a single unanticipated warm month or even a cold year (see Forum article) is not necessarily a sign that our understanding of long-term climate change trends is faulty. To understand and report responsibly on changes in Earth’s climate, the media need to probe scientific analyses of changes in the intensity of hurricanes or increases in annual temperatures occurring over the course of decades.
They should keep in mind too that inherent uncertainties in projecting short-term weather have little bearing on climatologists’ abilities to model long-term climate. They are not examining whether it will rain or not on a given day; rather, they are looking at how the rules governing precipitation rates, storm formation, and temperature will change in the long term.
In the end, meteorology and climatology are distinctly different fields using different methodologies. Journalists should keep those differences in mind when reporting on weather or on climate, which themselves are also two different things.
Meteorologists and Climatologists:
The Differences Matter
When should the news media turn to a climatologist to address questions? And when to a meteorologist? It’s a question reporters often ask each other.
Meteorologists study and attempt to forecast short-term changes in the weather lasting perhaps a few weeks. They work on improving those forecasts and on better understanding initial weather conditions so they can correctly model changes in weather over the course of days.
Climatologists, on the other hand, study the periodicity of weather events over years and millennia. They study changes in long-term average weather relative to atmospheric patterns to better understand local, regional, and global climates and natural and human-caused factors causing climate changes.
The disciplines are related, but sufficiently disparate that practitioners in one of the fields should not be assumed to be expert in the other. Just as climatologists cannot predict tomorrow’s weather with general circulation models, meteorologists might not understand all the elements that go into calculating long-term changes in Earth’s climate.
There certainly are some meteorologists with a deep understanding of climatology, but journalists should not assume that a meteorologist is necessarily qualified to comment on long-term climate change projections.