Climate Change Reconsidered II Physical Science Lead Authors/Editors

ion have primarily been responsible forchanges in the observed flows” of the negative-trendstations. Because Steynor et al. had presumed warmingwould lead to decreased flow, they assumed theirprojections were correct. Lloyd was able to demonstrate those results were driven primarily by unaccounted-for land use changes in the fivecatchments; the one site that had “a pristinewatershed” was the one with the “14% increase in flow over the study period.” Lloyd concluded hisresults were contrary to the climate changepredictions and indicate “climate change models cannot yet account for local climate change effects.” • Panin and Nefedov (2010) provided a riverine geomorphological and archaeological study of theUpper Volga and Zapadnaya Dvina Rivers (Russia) toassess the hypothesis that human settlement onfloodplains is controlled by the frequency of seasonalfloods. Their database comprised occupational layersfor 1,224 colonization epochs at 870 archaeologicalsites in river valleys and lake depressions insouthwestern Tver province.Panin and Nefedov identified a series ofalternating low-water (low levels of seasonal peaks,many-year periods without inundation of flood plains)and high-water (high spring floods, regular inundationof floodplains) intervals associated with periods ofwarming and cooling, respectively. The period AD1000–1300 Middle Ages provided particularlyfavorable conditions for floodplain settlement in areassubject to inundation today. Panin and Nefedovconclude this interval “can be regarded ashydrological analogues of the situation of the late20–early current century.” This relationship impliesthe current level of warmth in the portion of Russiathat hosts the Upper Volga and Zapadnaya DvinaRivers is not yet as great as it was during the AD1000–1300 portion of the Medieval Warm Period. • Zhang et al. (2010) analyzed twentieth centurydaily streamflow for eight unregulated streams in theSusquehanna River Basin, USA. This basin includesparts of the states of Maryland, New York, andPennsylvania and is the largest freshwater contributorto Chesapeake Bay, comprising 43 percent of thebay’s drainage area and providing 50 percent of itswater. The records studied start at slightly differenttimes, but all end in 2006 with record-lengths rangingfrom 68 to 93 years and an average length of 82.5years. The data were subjected to monotonic trendtests, each of which used different beginning andending times, to detect changes and trends in annualminimum, median, and maximum daily streamflow.Zhang et al. found annual maximum streamflow “does not show significant long-term change,” but there was “a considerable increase in annualminimum flow for most of the examined watersheds,and a noticeable increase in annual median flow for about half of the examined watersheds.” • Hannaford and Buys (2012) analyzed trends in U.K. river flow between 1969 and 2008 in a network of 89 near-natural catchments in an attempt todistinguish natural climate-driven trends from directanthropogenic disturbances. Previous model studies have suggested the U.K. will experience wetterwinters and hotter, drier summers in the future,causing decreasing river flow in summer and increases in winter (Murphy et al. 2009; Arnell, 2011; Prudhomme et al., 2012), with increases in floodfrequency and magnitude in some regions (Arnell, 2011; Kay and Jones, 2012; Bell et al., 2012). RealObservations: The Hydrosphere and Oceans 767world data, however, indicate droughts in 2004–2006(Marsh et al., 2007) and 2010–2012 (Marsh, 2012)were caused by successive dry winters, while asequence of wet summers occurred in the 2007–2012period (e.g., Marsh and Hannaford, 2008).In apparent harmony with climate modelprojections, Hannaford and Buys observed “an overallincrease in winter river flows.” But in conflict withwhat the models predict, they report “in summer,there is no compelling evidence for a decrease inoverall runoff or low flows, which is contrary totrajectories of most future projections.” Morespecifically, they found the predominance ofincreasing flow trends across the seasons, coupledwith limited decreases in low flows, is favorable froma water management perspective; the lack of anytendency toward decreasing river flow (for summerand for low flows especially) contradicts modelexpectations under assumed global warmingscenarios; and the lack of decreasing river flowindicates a robustness and resilience of hydrology towarming trends. • Khoi and Suetsugi (2012) evaluated seven climatemodels—CMIP3 GCMs-CCCMA CGCM3.1, CSIROMk30, IPSL CM4, MPI ECHAM5, NCAR CCSM3.0,UKMO HadGEM1, and UKMO Had CM3—todetermine which was most successful in predictingrates of streamflow in Vietnam’s Be RiverCatchment. The IPCC’s SRES emission scenariosA1B, A2, B1, and B2 were adopted, along withprescribed increases in global mean temperatureranging between 0.5 and 6°C.GCMs consistently have projected increases inthe frequency and magnitude of extreme climateevents, and variability of precipitation, leading some authors to conlude “this will affect terrestrial waterresources in the future, perhaps severely” (Srikanthanand McMahon, 2001; Xu and Singh, 2004; Chen et al., 2011). Khoi and Suetsugi’s findings, however,indicate “the greatest source of uncertainty in impactof climate change on streamflow is GCM structure [choice of GCM],” noting the range of uncertainty could have increased even further if a larger numberof GCMs had been deployed. Similar findings havebeen made by other authors, including Kingston and Taylor (2010), Kingston et al. (2011), Nobrega et al.(2011), Thorne (2011), and Xu et al. (2011). Khoi andSuetsugi conclude “single GCM or GCMs ensemble mean evaluations of climate change impact are unlikely to provide a representative depiction ofpossible future changes in streamflow.”ConclusionsThere appears to be little support in real-world datafor the contention that CO2-induced global warmingwill lead to more frequent and/or more severeincreases and decreases in streamflow that result in,or are indicative of, more frequent and/or more severefloods and droughts. Observed trends appear to bejust the opposite of what is predicted to occur, andnearly all observed real-world changes are either notdeleterious or are beneficial, and often extremely so.For example: • Lins and Slack (1999) report the United States hasbecome wetter in the mean and less variable in the extremes during warming over the last century. • Brown et al. (1999) have shown the 1999Mississippi floods were not related to changes in