As part of my GSoC 2026 project (#2750), I am working on adding support for diffuse irradiance components and component-specific optical loss calculations to ModelChain. This includes adding support for diffuse IAM models to ModelChain.
While planning this work, I noticed that all direct IAM models currently available in pvlib.iam are supported by ModelChain except for schlick. As noted by @kandersolar, there is no particular reason for this, it's just that no one's gotten around to it. So I would like to implement this.
This is relatively straightforward to implement, but I think there is one particular point worth discussing: whether schlick should be included in ModelChain's infer_aoi_model.
infer_aoi_model is used when the user does not provide an aoi_model, and selects an IAM model based on the parameters available in the PVSystem. The current priority order is: physical, sapm, ashrae, martin_ruiz, interp.
If included, schlick would presumably be in last place since it requires no parameters. This would also mean that the existing ValueError raised when no IAM model can be inferred would no longer occur, since schlick would always be available as a default. Therefore, this ValueError could be removed.
As part of my GSoC 2026 project (#2750), I am working on adding support for diffuse irradiance components and component-specific optical loss calculations to
ModelChain. This includes adding support for diffuse IAM models toModelChain.While planning this work, I noticed that all direct IAM models currently available in
pvlib.iamare supported byModelChainexcept forschlick. As noted by @kandersolar, there is no particular reason for this, it's just that no one's gotten around to it. So I would like to implement this.This is relatively straightforward to implement, but I think there is one particular point worth discussing: whether
schlickshould be included inModelChain'sinfer_aoi_model.infer_aoi_modelis used when the user does not provide anaoi_model, and selects an IAM model based on the parameters available in thePVSystem. The current priority order is:physical,sapm,ashrae,martin_ruiz,interp.If included,
schlickwould presumably be in last place since it requires no parameters. This would also mean that the existingValueErrorraised when no IAM model can be inferred would no longer occur, sinceschlickwould always be available as a default. Therefore, thisValueErrorcould be removed.