they say that actual observed data is unreliable so they'll use a made up model instead to prove their claims. A better paraphrase would be "they say that accu

they say that actual observed data is unreliable so they'll use a made up model instead to prove their claims. A better paraphrase would be "they say that accurate data on one part of the system are hard to get, so they compared several models using another, correlated, part of the system, for which better data are readily available." Also, since their claims relate to which of 16 models match "actual observed data" best, it's incorrect to characterize their work as using a model to prove their claims. The models are the subject of their claims, not tools used to prove them.