I didn't say human activity had caused GW, just that the evidence is clear that it has had a major impact. Approx 97% of climatologists accept this.
However you only get 97% if you have Diane Abbot type arithmetic skills.
The 97% figure come from work done by by Doran & Zimmerman of the University of Illinois at Chicago.
The used an online polling tool to ask 10,257 members of the American Geophysical Union (of whom 3,146 replied) a number of questions but there are only two that get talked about:
“When compared with pre-1800s levels, do you think that mean global temperatures have generally risen, fallen, or remained relatively constant?”
“Do you think human activity is a significant contributing factor in changing mean global temperatures?”
They whittled the 3,146 down to 77. Of them only 76 said 'risen' for Q1 and 75 said 'yes' to Q2. And 75/77 = 97.4%.
I don't think that 75/10,257 or even 75/3,146 is anything like 97%.
That totally leaves aside the fact that the second question is so vague as to be essentially meaningless. What is a 'significant contributing factor', 10%, 25%, 50%, 75%?
Also, in 1900 at least 97% of experts would have said that the continents were not moving, and they would have been wrong. In 1960 at least 97% of medics would have said that ulcers are caused by stress, and they would have been wrong.
Science is not a popularity contest. What matters is not how many people say something or how many papers are published on something. Scientists are people with, families, bills, etc - they will tend towards publishing work that fits the current funding vogue so that they can keep working. What matters is the facts.
A few key facts are that:
1. We have very little robust global data to work with.
The only close to global data is from satellites and that only goes back to the 1980s. Climate is normally defined as a 30 year average. So we have enough data to produce one reasonably reliable data point. Even that one data point has some caveats as some satellite instruments have gradually failed and satellite orbits have changed in ways that were not predicted which affected the way the results had to be interpreted.
2. Older data is very patchy. The oldest data set (Central England temperature) only goes back to 1659 or 1722 for the better data. In geological terms that is nothing and it only covers one part of England. Wider decent data (Europe, the British Empire, the US) only goes back to the 19th century.
3. Good ground level data is not available on a worldwide basis even today. Large chunks of Africa and Asia and both Polar regions are poorly covered.
4. Some of the ground level data from recent decades has questions about it due to the urban heat island. Does anyone think that Heathrow temperatures truly reflect London?
5. Atmospheric CO2 has gone up c. a third since 1950, see
http://woodfortrees.org/plot/esrl-co2/from:1900/to:2017 .
6. Global temperatures have increased but in a much less sustained fashion than CO2, see
http://woodfortrees.org/plot/uah5/from:1900/to:2017 .
7. Natural phenomena (e.g. the spikes on that graph which are the 1997-98 and 2014-16 El Nino events) are the biggest recent effects.
8. The late 20th century warming (late 70s to c. 1998) is almost identical to the early 20th century warming (mid 1900s to late 1930s), and the latter happened before there was a dramatic increase in atmospheric CO2.
My conclusion is that it is obvious that adding a significant amount of CO2 to the atmosphere is likely to have some effect on global temperatures. Some direct (by delaying heat loss to space) and some indirect (by increasing plant growth), and some of these may well act in opposite directions. What the overall impact will be no-one knows In the IPCC's First Assessment Report (1990) climate sensitivity (the change in global temperature for a doubling of CO2) was said to be most likely to be between 1.5 and 4.5 °C. Almost thirty years later that vitally important figure has not been nailed down any more accurately.
Reducing fossil fuel use (e.g. through more efficient appliances and insulation) has clear short-term benefits and seems obvious to me. Adding significant costs to fuel & taxpayer bills to subsidise unreliable wind & solar plants to possibly offset an unquantified possible future problem seems daft.