There are two methods can correct the GCMs by using observed data as I known.

You can refer the two publications for more details:

Article Observational constraints on Hyperinflation

Article Bias Correcting Climate Change Simulations - A Critical Review

Dear Nicholson,

There are several ways you can use to validate a climate model. For example, do a simulation for a certain region of 10 years, after this validation compare a model variable (eg Precipitation) with observed data from stations for the region where you made the simulation, and apply some statistical models to validate the model. Remembering this is a simple case. Below I leave a link to some papers that validate climate models.

https://link.springer.com/article/10.1007/s40641-016-0050-x

https://link.springer.com/article/10.1007/s00704-009-0134-9

https://link.springer.com/article/10.1007/s10584-019-02393-x

As an alternative to correcting and calibrating the individual GCM output, Multi-model ensembles is popular method to enhance the credibility of GCM predictions.

This method of Multi-model ensembles is usually used in studies related to projected climate scenarios; where several climate models are used together. The use of ensembles enables minimization of the uncertainty associated with individual model results.

Article Use of multi-model ensembles from global climate models for ...