Not when the issue is beyond the application of the scientific processes of verification and falsification because of such a multitude of interactions, positive and negative feed-backs, enormous complexity, elements of chaos, long time scales for a consistent trend to be observed, and relatively poor and inaccurate proxy records from the past with which to compare the present climate.
You're trying to wriggle out of it, but it still doesn't make sense.
With complex dynamics, it is customary to make models that allow testing of those interactions without having to wait for them to happen. The procedure/model is peer reviewed and if almost all others arrive at the same conclusions, and find that the model is adequate, there emerges a consensus about the findings.
As I mentioned before, the Technical summary of the latest IPCC report stated there was low confidence that floods, droughts and storms had been increasing during the previous 50 years or so, on a global scale. Floods, droughts and storms are major events, yet climate science observations are not even able to determine with confidence whether or not they have been increasing in recent times. What does that tell you?
Exactly what they say in the report, that the (historical) data is lacking in some respects (e.g. measuring location had to be moved due to rising water levels, or urbanisation coming too close to avoid interference, or no prior measuring point available but now a new point has been made available for improved coverage and data quality going forward). Sometimes a new method of data collection is replacing a less accurate method. So there is no longer an exact 1:1 relationship between historical measuring locations and new measuring locations in the same area. That's all.
It seems like I still have to explain it, "confidence levels," tell something about the data quality and consensus that follows from those observations. You are presumably still thinking that it is the same as a likelihood, but that's something else (i.e. the probability of something happening).
The confidence level can be increased by e.g. more frequent observations or more accurate observations, or spatially more dense observations. This becomes easier now than it was in the past, because the quality of sensors is increasing, and the cost is often dropping, thus allowing to deploy more of them (where few or none were there before). The larger quantities also allow to more easily identify outliers, and its data can then be validated or rejected before being added to the final dataset.
Cheers,
Bart