2.1 Climate Basics#
2.1.1 What is Climate?#
A common point of confusion for people new to climate science is the distinction between climate and weather. An excellent idiom to explain the difference is that the weather determines the clothes you wear today, but the climate determines the type of clothes you keep in your wardrobe. To be more precise, weather is a short timescale phenomenon. Weather is day-to-day, hour-to-hour, and even sub-hourly variations in the state of the earth system. The climate of a particular region is determined by long-term averages of weather conditions.
Another important distinction between weather and climate is that weather is chaotic. You probably know that weather forecasts can only be made for up to at most two weeks into the future, and become more accurate as the lead time gets shorter. This is because weather prediction is mostly an initial value problem, and the highly complex nature of the earth system means that the sequencing of weather patterns is extremely sensitive to the initial conditions. Since the initial conditions for weather forecasts ultimately come from observations, which are subject to instrumental limitations in both precision and spatiotemporal coverage, there is always some amount of error between the initial conditions for a forecast and the true state of the earth system at that point in time. This error grows in magnitude over time as a forecast is produced, and at around the two-week mark the output of forecast systems becomes completely decorrelated with reality. This is the reason why weather forecasts are only useful for rather short time periods.
Fortunately, modeling of the climate is not subject to this same limitation. Because climate is dictated by long-term tendencies, the particular sequencing of weather patterns (the entire purpose of weather forecasting) is irrelevant. Climate prediction is what we call a boundary value problem - the results are dictated not by the starting point, but by the external conditions you apply. These include things like the amount of incoming solar radiation, concentrations of greenhouse gases and other substances in the atmosphere, land use and land cover, and others. All of these forcings are included in weather prediction models, but they are not the limiting factor in the accuracy of the forecasts. For climate modeling, these forcings are critical for producing a realistic climate state, and changes to these forcings to reflect plausible human actions are how climate change emerges in the models.
2.1.1.1 The Climate System#
The climate (as defined above) is produced through interactions between several physical systems, which together are called the climate system. The glossary of the Intergovernmental Panel on Climate Change (IPCC)’s 6th Assessment Report defines the climate system as “The global system consisting of five major components: the atmosphere, the hydrosphere, the cryosphere, the lithosphere and the biosphere and the interactions between them. The climate system changes in time under the influence of its own internal dynamics and because of external forcings such as volcanic eruptions, solar variations, orbital forcing, and anthropogenic forcings such as the changing composition of the atmosphere and land-use change.” The second sentence of this definition helps us understand factors that result in variations in the climate, especially those not due to human influence, i.e. natural climate variability.
2.1.2 Natural Climate Variability#
The climate system varies over a number of different characteristic timescales. Tropical regions are warmer than polar regions - this is easily determined by comparing multi-annual means over all days of the year. But as we intuitively know, the mean climate varies predictably and consistently on sub-annual timescales through the seasonal cycle. In the Northern Hemisphere winter, the tilt and position of the Earth relative to the sun means this region receives less solar radiation than in the summer when the Earth’s rotation and orbital position favours the absorption of more sunlight. For this reason, the average temperatures of the winter months are colder than for the summer months. Tropical regions may not experience the same drastic swing of typical temperatures from one season to the next, but instead, they may experience a seasonal cycle in the amount of precipitation between wet and dry seasons. We can define for each month, or for each day of the year, the climatological mean as the long-term average conditions for that month/day of the year. We can also define for each of these time units, a typical range of conditions using a measure of spread (such as the standard deviation or interquartile range) calculated across many years. One can combine these two measures to give a more complete description of the climate for a given region, as the overall spread of typical weather conditions, and how this range of conditions typically evolves over the course of a year.
The seasonal cycle is a source of natural climate variability that nearly everyone is familiar with from their own interaction with the outdoors. There are of course many other sources of natural climate variability, many of which are substantially less predictable than the seasonal cycle. The climate system can vary interannually (year-to-year) due to volcanic eruptions (which cool the Earth by blocking sunlight) or periodic phenomena such as the El Nino-Southern Oscillation (ENSO). The climate can also vary on decadal timescales due to periodicity in the intensity of the Sun’s radiation, and over centuries or millennia due to variations in the Earth’s orbit and axis of rotation.
Variations such as those due to ENSO are due to dynamics internal to the climate system and are not caused by external factors such as solar forcing. Thus, it is referred to as internal variability, which is a subset of natural climate variability (since solar cycles are still natural). Like weather, internal variability is chaotic and less predictable than the effects of external forcing like solar cycles or greenhouse gases. For the same external forcing, there can be many possible realizations of internal variability, and it’s important to remember that we’re only living through one of them. Sometimes what may appear to be a real trend in observations can simply be a result of random fluctuations due to internal variability. In models, we can make good attempts to study changes to the climate in the context of internal variability, but further discussion is left for Section 2.3.
This list of sources of natural climate variability is not even close to exhaustive, but it illustrates that climate is a moving target, even in the absence of any human influence. When studying anthropogenic climate change, we need to be able to separate the effects of human influence (like increased greenhouse gas concentrations) from the effects of natural variations in the climate, and check whether changes we think we see in the data truly rise above historical natural variability.