Revisiting: What is climate?

Climate is what we expect, weather is what we get.

- Mark Twain

What is climate? Is it different from weather?

Before considering the "when" of climate, let's look at the "what" and "where." Is climate really just about the atmosphere? Not really - all the funny things water does in the atmosphere means that "weather" and "climate," at a minimum, must include the sphere of air and the sphere of water - call it the "aerohydrosphere."

Climate and weather refer minimally to the aerohydrosphere.

Just as we determined almost a year ago that local climate state required, not just temperature, but wind, water state, and humidity as well, so we conclude that "climate" defined in substance and space must include both aerosphere and hydrosphere.

But we also have learned over the last year that climate is affected by the action of plants and the interaction of geochemistry with the oceans. (These have particular importance when we're thinking about the evolution of carbon dioxide in the climate system.) When referring to these, we need to expand "climate" to include "bio" and "geo" parts.

Climate, as an extended and more complete system, is the state of the aerohydrobiogeosphere.

And in postings to come, we'll learn that the Earth's climate is affected in important ways by changes in the Sun. They form a coupled helio-geosystem.

Climate, in its most complete definition, is the state of the aerohydrobiogeoheliosphere.

Our intuitions about climate and weather are deeply tied to differences in time scale. Instead of abstract definitions, start by clarifying the ordinary language. By "weather," most people mean the changeable atmospheric state over periods of hours or days or weeks.* By "climate," most people mean the atmospheric state over months or years or decades - or perhaps longer. We run into the limits of the human life span, and that biases us into ignoring even longer time scales. But climate includes them as well.

This intuition about "climate" versus "weather" is sound enough, although it's fuzzy and needs to be made more exact. Many weather phenomena are periodic or quasi-periodic, and it's often best to consider "climate" as happening on time scales longer than the longest internal "natural" period of the system. Earth climate has only two exact periods imposed on it by external cycles, those of the day and of the year. They are imposed by astronomical cycles having fundamentally nothing to do with climate: the rotation of the Earth on its polar axis and its orbit around the Sun, respectively.

It will be helpful to define "climate," with respect to time scale of change, as:

Aerohydrospheric change over time scales longer than a year.

Anything on shorter time scales is "weather."

Now consider more carefully the issue of cycles. Besides the cycles imposed from the outside by astronomical behavior, the climate has internal periodic modes, with periods extending from months to centuries or millennia. There are also modes that flare up and die away.**

It is this sort of behavior that leads some people to believe, wrongly, that climate has a steady state, one arrived at by averaging away all periodic behaviors and waiting for transient "dying-away" phenomena to "settle down." This is a convenient conceptual picture of climate for theoretical purposes. The transport of radiation and water, in particular, can be smoothed and averaged in this way, leading to a climate steady state of steady clouds, steady evaporation, steady precipitation, steady radiative light flux in and heat flux out, and steady convection upwards.

But it is NOT the real climate - by design or by accident, we've excluded the hard part, the tricky piece due to chaos. In the atmosphere and oceans, it appears in the form of fluid turbulence, part of the larger phenomenon of convection. Such chaotic systems exhibit some behavior that's periodic, some that's "dying away" (transient), but a third behavior as well - one that never dies away, but is also not periodic. Chaos consists of a stream of unique, never-to-be-repeated events. It is the shadow that falls on all our dreams of predictability and control, the metaphysical fly in the ointment.

We can arrive at a "steady state" picture of climate only by deleting the reality of turbulence-chaos, then averaging away all other change. Real weather unavoidably is threaded by unique discrete events that prevent anyone from taking any climate as "steady" or "typical."

When we talk about an idealized climate as "steady," "typical," or lacking a unique history, we have implicitly removed turbulence-chaos. It's a useful conceptual construct. But it's not the climate we live in, the one visible outside, the unsteady parade of the unpredictable weather. The most serious mistakes made with climate, very much including mistakes by people who should know better, start here.

Perhaps a better way to state the difference between "climate" and "weather" is this:

Climate is weather with the chaos removed. Weather is climate with the chaos put back in.

---
* Not coincidentally, weather is predictable up to about two weeks ahead. Most people know this, just not in their frontal lobes.

** Dying away exponentially or by some inverse power of time.