This latest report by Paul Homewood is most interesting. The mainstream media, since the recent record rainfall and flooding in and around Baton Rouge, LA, had a field day with aspects of the story. Instead of focusing exclusively on the fatalities, property damage, and utter destitution of the thousands of flood victims, the major media decided to make hay, so to speak, with claims first arising from some climate-change alarmists that global warming is responsible for rainfall events. These events are alleged to become more frequent, and more severe, than in the past as a result of CO2-induced climate change.
But a carefully designed statistical study of rainfall patterns in the United States using coupled models and input data obtained from both moderate and finely defined grids (where available) did not show an acceptable level of predictability or significant departure from established norms within the range of natural variability.
Environmental journalists who work for several leading media outlets such as the AP and the New York Times seem to have difficulty resisting the temptation of executing the principle of “ready, fire, aim” when it comes to analyzing these weather disasters. It also appears their editors are reluctant to exercise appropriate oversight, although outside observers can’t be sure that similar stories have not been spiked.
Also noteworthy is Mr. Homewood’s concluding observation:
“As we often see, the computer models show more intense rainfall, because that is what they are programmed to say. As so often though, when the observed record is analyzed, no evidence of such is found…”
His conclusion comes as no surprise to those aware of how the elaborate climate models are constructed. Built into the models are certain key assumptions and mathematical procedures guaranteed to project warming temperatures and increasing precipitation with increasing greenhouse gas concentrations. In that regard they incorporate an unavoidable, built-in element of circular reasoning.
From Paul Homewood’s Blog:
September 10, 2016
By Paul Homewood
This paper, published last month, has some relevance to recent attempts to blame the Louisiana floods on global warming.
Precipitation extremes have a widespread impact on societies and ecosystems; it is therefore important to understand current and future patterns of extreme precipitation. Here, a set of new global coupled climate models with varying atmospheric resolution has been used to investigate the ability of these models to reproduce observed patterns of precipitation extremes and to investigate changes in these extremes in response to increased atmospheric CO2 concentrations. The atmospheric resolution was increased from 2°×2° grid cells (typical resolution in the CMIP5 archive) to 0.25°×.25° (tropical cyclone-permitting). Analysis has been confined to the contiguous United States (CONUS). It is shown that, for these models, integrating at higher atmospheric resolution improves all aspects of simulated extreme precipitation: spatial patterns, intensities and seasonal timing. In response to 2×CO2 concentrations, all models show a mean intensification of precipitation rates during extreme events of approximately 3-4% K−1. However, projected regional patterns of changes in extremes are dependent on model resolution. For example, the highest-resolution models show increased precipitation rates during extreme events in the hurricane season in the CONUS southeast, this increase is not found in the low-resolution model. These results emphasize that, for the study of extreme precipitation there is a minimum model resolution that is needed to capture the weather phenomena generating the extremes. Finally, the observed record and historical model experiments were used to investigate changes in the recent past. In part because of large intrinsic variability, no evidence was found for changes in extreme precipitation attributable to climate change in the available observed record.
As we often see, the computer models show more intense rainfall, because that is what they are programmed to say. As so often though, when the observed record is analysed, no evidence of such is found:
Finally, the observed record and historical model experiments were used to investigate changes in the recent past. In part because of large intrinsic variability, no evidence was found for changes in extreme precipitation attributable to climate change in the available observed record.