In general, there are only two way to prove something in science.
1. Prove a singular (fact) with an observation such as “black swans exist”.
2. Disprove a universal (theory) with a singular fact such as “all swans are white”.
The inability to disprove a singular, or to prove a universal, is due to our finite limits to our observations. In general, we cannot gather the infinite observations required disprove (1) a fact, or to prove a universal (2).
Scientists need to be rigorous and strict particularly in the initial stages of formulating a study, whether it is a singular or a universal that is being tested, and how the observations will impact.
A case in point: the impact of observations of global temperatures on the climate model projections plotted below. By a strict interpretation of scientific method, the observed “slow rise in global temperature” is a fact that disproves the universal “all possible trajectories of climate models under AGW warming”.
The only appropriate scientific response is to throw away all of those falsified models and all of the work based on them – extinction predictions, extreme events, agricultural trends, and so on – as it is scientifically worthless. You must go back to the drawing board.
The rules of science were illustrated recently in a post on Vortex about the Wright Brothers’ first flight:
To give another dramatic example, suppose at 1:00 pm on the afternoon of December 17, 1903, you were take a poll about whether man can fly. Suppose you asked people to place bets as to whether airplanes exist. Out of the 1.6 billion people in the world alive on that day, at that moment, the only ones who had ANY KNOWLEDGE of that question were Wilbur and Orville Wright and the members of the Kitty Hawk coast guard who had helped them fly that morning. In all the world, there was not another soul who knew the facts or was qualified to address the question. The opinions of other people were worthless. Meaningless. All the money in the world placed in a bet would mean nothing. There was an undeveloped glass plate photograph showing the first flight:
That photograph was proof. It overruled all opinions, all money, all textbooks, and the previous 200,000 years of human technology. A thermocouple reading from a cold fusion experiment in 1989 overrules every member of the human race, including every scientist. Once experiments are replicated at high signal to noise ratios, all bets are off. The issue is settled forever. There is no appeal, and it makes no difference how many people disagree, or how many fail to understand calorimetry or the laws of thermodynamics. The rules of science in such clear-cut cases are objective and the proof is as indisputable as that photograph.