3 ms·

Your comment doesn't match what the paper actually says[1]. Your claim that "The old studies didn't predict how much CO2 the world would emit" is directly cont

Your comment doesn't match what the paper actually says[1]. Your claim that "The old studies didn't predict how much CO2 the world would emit" is directly contradicted on the first page. The paper begins by pointing out that "model projections rely on...accurate assumptions around future emissions of CO2". In fact, one of the criteria the authors used for choosing a model to include was that it came with a time-series prediction of "future projected atmospheric CO2 concentrations". With that restriction, the authors found 11 papers containing a total of 14 models. Unless the authors of this paper left out other predictions from the original papers, the only papers included in the study that "predicted how much the planet would warm, for various amounts of CO2 emissions" were Hansen's, and the other 9 papers included in this study only made one prediction for both emissions and temperature. This paper also points out that Hansen's 1988 "most plausible" model, which you described as "the one model that was presented to Congress", was not accurate under the usual definition of accurate: "H88's “most plausible” scenario B overestimated warming experienced subsequent to publication by around 54%." Then how does this paper lead to the headline that "Even 50-year-old climate models correctly predicted global warming"? What the authors of this paper have done is to compare errors in the models' predictions of CO2 emissions to errors in the models' predictions of temperature. In other words, the models were wrong about both, but in the same direction. They overestimated both future emissions and future temperature. Even so, the claim that the models were accurate is a linguistic trick. The table on line 326 shows what the authors are calling skill score: "A skill score of one represents perfect agreement between a model projection and observations." Not a single model has a "skill score" of one, or for which one is within the uncertainty. In fact, all of the models are wrong in the same direction, even after this paper has attempted to correct for their incorrect assumptions about future CO2 emissions. The linguistic trick is that this error is not, according to the authors, "statistically significant". But that's a tremendous abuse of the phrase. The authors are effectively saying: "The null hypothesis is that the models are right, and we've failed to disprove it." They've assumed the conclusion. I suspect that this paper is being promoted for political, not scientific, correctness. 1: A PDF of this paper can be found at https://pubs.giss.nasa.gov/docs/tbp/inp_Hausfather_ha08910q.pdf https://pubs.giss.nasa.gov/docs/tbp/inp_Hausfather_ha08910q....

DennisP 7y ago
All that says (in your second paragraph) is that if you're going to predict the actual temperature, you need an accurate CO2 prediction, which of course is true. But here is Hansen's original 1988 paper: https://pubs.giss.nasa.gov/abs/ha02700w.html https://pubs.giss.nasa.gov/abs/ha02700w.html From the abstract: "We make a 100-year control run and perform experiments for three scenarios of atmospheric composition. These experiments begin in 1958 and include measured or estimated changes in atmospheric CO2, CH4, H2O, chlorofluorocarbons (CFCs) and stratospheric aerosols for the period from 1958 to the present. Scenario A assumes continued exponential trace gas growth, scenario B assumes a reduced linear linear growth of trace gases, and scenario C assumes a rapid curtailment of trace gas emissions such that the net climate forcing ceases to increase after the year 2000." Note the word "assumes." There's no attempt to really model what humanity would choose to emit. None of these assumptions exactly match what happened with CO2 and other greenhouse gases, so of course none of them exactly match the temperature either. But adjusted for actual greenhouse emissions, they come remarkably close. Were they completely accurate in that way? Certainly not. People knew a lot less back then. It took a while to nail down things like the effect of water vapor, which is a greenhouse gas but also makes clouds that reflect sunlight away. But they were close enough to make good policy decisions, and if we'd paid attention, we'd be much better off today. Incidentally, the case for global warming doesn't depend on models. There have been warming events before, and there's enough evidence in geological history to tell us what's going to happen. Hansen's book has details.
- leereeves 7y ago
  > But adjusted for the actual CO2 emissions, they come remarkably close. Take a look at the table I mentioned (line 326), in particular "ΔT / ΔF skill". That's the change in temperature over the change in forcing (ie, CO2 etc). Hausfather et al call it "implied TCR". In (overly) simple terms, it's the predicted increase in temperature per unit of CO2 emitted. Look at the numbers: 0.51, 0.41, 0.63, 0.42, 0.83, etc. The best is 0.87. The average was 0.69. The authors attempted to adjust for actual CO2 emissions and found that, after the correction, the models were predicting 40%-90% of the observations, with an average score around 70%. Is that "remarkably close"? Well, we might disagree about the meaning of that term. But there's another observation that we should agree on. Note what the authors say after that table: > The average of the median skill scores across all the model projections evaluated is 0.69 for the temperature vs time metric. > Using the implied TCR metric, the average projection skill of the models was also 0.69. In other words, the models are, on average, as accurate with or without the correction. Adjusting for actual greenhouse emissions doesn't make the models more accurate (nor less).

DennisP 7y ago

All that says (in your second paragraph) is that if you're going to predict the actual temperature, you need an accurate CO2 prediction, which of course is true. But here is Hansen's original 1988 paper: https://pubs.giss.nasa.gov/abs/ha02700w.html https://pubs.giss.nasa.gov/abs/ha02700w.html From the abstract: "We make a 100-year control run and perform experiments for three scenarios of atmospheric composition. These experiments begin in 1958 and include measured or estimated changes in atmospheric CO2, CH4, H2O, chlorofluorocarbons (CFCs) and stratospheric aerosols for the period from 1958 to the present. Scenario A assumes continued exponential trace gas growth, scenario B assumes a reduced linear linear growth of trace gases, and scenario C assumes a rapid curtailment of trace gas emissions such that the net climate forcing ceases to increase after the year 2000." Note the word "assumes." There's no attempt to really model what humanity would choose to emit. None of these assumptions exactly match what happened with CO2 and other greenhouse gases, so of course none of them exactly match the temperature either. But adjusted for actual greenhouse emissions, they come remarkably close. Were they completely accurate in that way? Certainly not. People knew a lot less back then. It took a while to nail down things like the effect of water vapor, which is a greenhouse gas but also makes clouds that reflect sunlight away. But they were close enough to make good policy decisions, and if we'd paid attention, we'd be much better off today. Incidentally, the case for global warming doesn't depend on models. There have been warming events before, and there's enough evidence in geological history to tell us what's going to happen. Hansen's book has details.

leereeves 7y ago
> But adjusted for the actual CO2 emissions, they come remarkably close. Take a look at the table I mentioned (line 326), in particular "ΔT / ΔF skill". That's the change in temperature over the change in forcing (ie, CO2 etc). Hausfather et al call it "implied TCR". In (overly) simple terms, it's the predicted increase in temperature per unit of CO2 emitted. Look at the numbers: 0.51, 0.41, 0.63, 0.42, 0.83, etc. The best is 0.87. The average was 0.69. The authors attempted to adjust for actual CO2 emissions and found that, after the correction, the models were predicting 40%-90% of the observations, with an average score around 70%. Is that "remarkably close"? Well, we might disagree about the meaning of that term. But there's another observation that we should agree on. Note what the authors say after that table: > The average of the median skill scores across all the model projections evaluated is 0.69 for the temperature vs time metric. > Using the implied TCR metric, the average projection skill of the models was also 0.69. In other words, the models are, on average, as accurate with or without the correction. Adjusting for actual greenhouse emissions doesn't make the models more accurate (nor less).