Overview on decadal prediction

The overarching goal of MiKlip is to develop a decadal climate prediction system. These predictions cover a time period from 1 to 10 years and close the gap between short-term weather forecasts, seasonal climate forecasts and long-term climate projections, which are today used in a wide range of applications such as short-term planning for when to harvest, in the long-term planning for environmental protection and climate adaptation.

Predictability of the climate system arises from certain processes that can be predictable, if the initial conditions (from observational data) and the physical laws describing the earth system (from theory) are known. The various components of the climate system (atmosphere, ocean, land and cryosphere) react with different delay to external forcings and thereby show “memory” for climate processes on different time-scales. Thereby, the atmosphere has a “memory” of days, sea ice and the land surface of 1-2 years and the ocean of 1-100 years. The “memory” of these components allow for predictability of climate variables on the decadal time-scale.

The bottom part of this graphic shows the time-horizon of different weather and climate forecasts and projections. The upper part indicates whether initial and/or boundary conditions provide the information important for the predictability of these.

It is not possible to use the decadal climate prediction system to predict the weather of the coming ten years for a particular day in a particular area. The aim is rather to predict climate tendencies over longer time-periods (e.g., 4-year means) and larger areas (e.g., 250-500 km) and to communicate these as deviations from a pre-defined normal state.

In order to learn more about decadal climate predictions in general, visit one of the four buttons on the right.

The MiKlip decadal climate prediction system is currently in development and all forecasts should be seen as experimental. Decadal climate predictions should not be used as the sole basis for decisions.