Weather and climate
There’s a common saying when it comes to describing the difference between weather and climate: “Climate is what you expect, weather is what you get.” Surprisingly, American author Robert A. Heinlein, rather than a scientist, is responsible for this inspired catchphrase.
Meteorology refers to the description and prediction of the state of the atmosphere in the short term (for periods ranging from a few hours to about ten days) through variables such as temperature, type and amount of precipitation, atmospheric pressure, humidity, cloud cover and sunshine. Weather forecasting models attempt to predict what the state of the atmosphere will be at a given time and place with the greatest possible accuracy. A successful forecast is one that comes true at the time and place predicted.
To produce a forecast, a weather forecasting model must initiate a short-term simulation based on the most accurate initial conditions possible. These are provided to the model in the form of an analysis generated by a system that assimilates observed data from the atmosphere, the oceans and land surfaces.
Climatology is concerned with the statistical characteristics of the same variables over at least three decades. By using multi-decadal series of values representing the changes in meteorological variables over time, it’s possible to calculate the average and variability of meteorological variables in order to determine the seasonal “normals,” which are the 30-year average of each variable for a given region. In addition to normals and variability, climate is also defined by the return periods of extreme events, as well as historical records such as the highest maximum temperature ever recorded for a given day or the largest accumulation of snow in 24 hours.
To study the past climate (or historical climate), the multi-decadal data sets required may come from observations or from long-term simulations by climate models. However, the study of the future climate relies exclusively on data simulated by climate models.
The success of a climate simulation of the past (like the modern historical period 1995-2014 used in the Sixth IPCC Report, or the period 1990-2020 used by the WMO) does not indicate its ability to reproduce the precise sequence of weather events observed in the past, but rather to generate statistical characteristics, such as the normals, variability and trends, that are similar to those observed.
The study of the future climate is based on a prospective rather than a predictive approach. For a future period (e.g. 2041-2070), it considers a plausible scenario of the state of the world population and its socio-economic conditions, which is translated into a greenhouse gas emissions scenario. A climate simulation based on a future greenhouse gases scenario is a climate projection. Its purpose is not to predict the weather in a specific place on a given day, but to see whether the greenhouse gases described in the emission scenario will affect the normals, variability or trends of the various meteorological variables in a region over a given time period. For example, we may forecast the weather for next weekend, but we make projections of the future climate in the middle or end of the century.
The fact that these two disciplines are so closely related contributes to the confusion. The phenomena that influence weather and climate are the same, as are the mathematical equations used in their modelling. Over the years, weather forecasting models and climate models have become more and more similar. However, they are not run in the same way or in the same context. For example, weather models need the observed state of the atmosphere provided by an analysis, which is not required for climate models.
This is why the distinction between weather and climate is not always clear to non-specialists. For example, during snowstorms, we often hear about harsh climate conditions when we should be talking about harsh weather conditions.
That said, there are many opportunities to combine weather and climate information. For example, when an official weather report explicitly refers to cold weather, it means that the temperature recorded or forecast in the coming hours is below normal for the season and deviates from it beyond the interannual variability of the region concerned. This means that the cold weather has nothing to do with our subjective perception, but instead with the relationship between the recorded or predicted temperature and the historical climate characteristics of the region established through 30 years of observations.
