If you're talking about whether theories match the real world, then you're correct. It can never be conclusively verified due to the problem of measurement.
However, given a theory, you can prove whether or not something is correct in the model described by the theory. So while it can never be conclusively proven whether, say, the standard model matches the real world perfectly, you can most definitely prove that electrons behave in a specific way in a world that behaves like the standard model says it should.
That leaves open the question whether the real world matches the standard model or not, and while very compelling measurements and experiments can be produced, we don't really know. Well, we do know, actually. There's no gravity in the standard model, so it most definitely does not describe our world, which is why the whole string theory research is happening.
But climate science does not have that. It does not have tiny theories with constants that try to find the smallest possible unit and how it behaves. Climate science laws talk about volumes of air, say cubes of 10km on a side, and statistically it will behave like X. But nobody really knows why we got X, except that it "tends" to predict measurements fairly accurately. Why ? We have some idea, like "co2 keeps sunlight from reflecting out", but we don't know why the observed concentrations in the athmosphere don't match what you see in a gas experiment (basically in experiments greenhouse effects stop at ~250-280ppm depending on concentrations of other gases, doesn't happen in the athmosphere obviously). But climate science doesn't even try to explain this, or reduce it to first principles, they just try to model whatever they see.
However, given a theory, you can prove whether or not something is correct in the model described by the theory. So while it can never be conclusively proven whether, say, the standard model matches the real world perfectly, you can most definitely prove that electrons behave in a specific way in a world that behaves like the standard model says it should.
That leaves open the question whether the real world matches the standard model or not, and while very compelling measurements and experiments can be produced, we don't really know. Well, we do know, actually. There's no gravity in the standard model, so it most definitely does not describe our world, which is why the whole string theory research is happening.
But climate science does not have that. It does not have tiny theories with constants that try to find the smallest possible unit and how it behaves. Climate science laws talk about volumes of air, say cubes of 10km on a side, and statistically it will behave like X. But nobody really knows why we got X, except that it "tends" to predict measurements fairly accurately. Why ? We have some idea, like "co2 keeps sunlight from reflecting out", but we don't know why the observed concentrations in the athmosphere don't match what you see in a gas experiment (basically in experiments greenhouse effects stop at ~250-280ppm depending on concentrations of other gases, doesn't happen in the athmosphere obviously). But climate science doesn't even try to explain this, or reduce it to first principles, they just try to model whatever they see.