Background
The author reports that "it is generally assumed that various environmental changes, including hydrological ones, can be anticipated as logical consequences of ... predicted climate change," and that "these predictions are presumed to provide the rock-solid foundation for policies to cope with the change."  It is his feeling, however, that "these methods of international global change science are highly skewed toward a rather particular set of philosophical presumptions," and that "a more appropriate organization to the global change research programme would not only make more effective use of palaeoflood hydrology, it would also much more optimally benefit humankind."

What was done
In an analysis of this philosophical dichotomy, Baker compares the two different ways of assessing future flood risks.  The approach that is favored by the IPCC uses what he calls "idealized conceptual models," which predict future flood behavior on the basis of global climate model (GCM) predictions of CO2-induced global warming and how that warming is projected to alter the planet's hydrologic cycle.  The other approach seeks to determine how real-world flooding in ages past responded to different global temperature regimes, and it assesses future flood risks on the assumption that floods yet to come will likely reflect the consequences of a similar hydrologic response to projected changes in the thermal regime of the planet.

What was learned
With respect to the first of these approaches, Baker reports that it leads to "the claim that the magnitudes of extreme floods may be increasing because of global warming."  With respect to the second, he notes that a preliminary survey of historical and palaeoflood records suggests that "the floods of recent years do not generally exceed ... those of past clusters, and much larger floods are usually indicated in the past," when climate alarmists claim it was much colder than it is now.  Consequently, the two approaches currently produce diametrically opposed predictions.

What it means
So what is one to believe?  Does one accept predictions about the future that are contrary to relationships known to have existed in the past?  Or does one go with the method that has a proven track record?  Baker concludes that "for 20 years there has been an overemphasis on developing [GCMs] as the principal tool for dealing with threats to habitability of the planet," stating that "in contrast to unverified predictions of the future given by GCMs, palaeoflood data provide evidence of real-world cataclysms that people can understand sufficiently to alter their perceptions of hazards."  And as we read the palaeorecord (see Floods in our Subject Index), it would appear that flooding will become less extreme than it is now if the planet continues to warm.