What are climate scenarios?

Climate scenarios present information on how the climate could possibly develop during a determined period of time in the future. They are created by government authorities dealing with meteorology or climate issues and/or research institutes with the help of global climate models. The calculations are based on assumptions about future changes in the atmosphere. They include the relationships between physical processes in the entire atmosphere-land-water system, as well as emission scenarios, which are assumptions about future emissions of greenhouse gases. Results from these global models can be downscaled further with regional models to provide greater detail. 

The results which are presented from calculations with the climate models are called climate scenarios. Climate scenarios are not weather forecasts. Weather forecasts give information about what is probably going to happen at the local scale during a short period in time. Climate scenarios represent the statistical behavior of weather, which we call climate, but they do not recreate the real weather at a specific place and time. 

When comparing future climate to our current climate, reference periods are often used e.g. of 20 or 30 years. The results for the future are often compared with the average for this reference period. Normally the period 1961—1990 is used as standard reference.

As the information from climate scenarios are calculated in the form of a grid (gridded data), it is difficult to compare this information to current climate observations, which are taken at specific locations (point data). Observations describe the conditions at a specific point, while models describe conditions evenly distributed over the whole grid. One can look at the example of precipitation. If a large amount of rain falls over a very small area, it would be recorded at one measuring station. At other nearby places which might have only received small amounts of rain or no rain at all, the measurement stations will record much smaller amounts. If the same total amount of precipitation is created in the model for the grid in question it will be evenly spread over the grid. This would make it appear as if it has rained evenly over the entire the grid while in reality (compared to observational measurements) the rain was much more unevenly distributed.