To validate whether or not climate models are providing us with accurate information about the Earth System, outcomes from model simulations are compared to observational data to assess where they are similar and where they diverge. Climate data, including both present day and historical data, is available and accessible. If there are significant differences, it means model developers have more work to do and must go back, check their data, check their algorithms, and see if they are missing any critical components of the system.
COMPARISONS AMONG CLIMATE RESEARCH GROUPS
Different research groups will come up with different mathematical equations to represent the same climate process. More and more, different climate modeling research groups are comparing their different methods to each other to see which ways best match what is observed in the climate.
Because scientists are able to compare their climate models to both historical and present day data, they are able to validate and improve their efforts over both the long term and the short term. Models are improving all the time. Here are a couple of examples from climate models that show particularly good comparisons with observations for specific weather events:
|1.5km resolution simulations of Typhhon Megi (Courtesy UK Met Office)|
|Meteostat satellite observations of clouds (left) are compared to climate model output from ECMWF (right)|
Validating climate models requires taking accurate, consistent observations over many decades, which is achieved through sustained international collaborations. This topic is covered in more depth in the recent report on A National Strategy for Advancing Climate Modeling.