A group of scientists in the UK Met Office has proposed replacing the current method of calculating temperature trends from 30 years of actual measurements — set almost 100 years ago by the International Meteorological Organization. Instead, Met Office scientists want to use 10 years of actual data combined with a 10-year model projection. In a Nature paper they point out that the Paris Agreement contains no formally agreed way of defining the present level of global warming, and without an agreed metric there can be no consensus on when the agreement's 1.5°C target has been exceeded. Even reducing the averaging period to 20 years of observed data means that a breach of the 1.5°C limit around 2030 wouldn't be recognized until around 2040, an unwelcome delay of 10 years.

To get around this delay the scientists propose a new indicator — a 20-year average temperature rise centered around the current year by blending observations for the past 10 years with model projections for the next 10 years and then averaging the two periods. They claim this method would provide an "instantaneous" indicator of current warming.

Chris Morrison, the Daily Sceptic's environment editor, notes that the Nature paper's authors used a computer model pathway of RCP4.5 that allows for a possible rise in temperatures of up to 3.2°C within 80 years. (Over the past 25 years observed warming has been about 0.2°C). He notes the poor record of climate models in producing accurate temperature forecasts and lack of confidence in global surface temperature records.

Credits to Friends of Science
friendsofscience.org