WasthecirqueglaciationofWalestime-transgressive,ornot? Ian S. Evans Department of Geography, University of Durham, Durham City DH1 3LE, England ABSTRACT . The lowest glacial cirques inWales are in the SouthWales Coalfield and inwesternmid-Wales:thehighestareinthenortheastandonthehighestmountains.Floor altitudes show great local variability, but in general rise to the northeast in most ofWales and northward in southernWales, as does the former glaciation level from 470^710m a.s.l.).The pattern is similar for reconstructedYounger Dryas glaciers, which occupied the higher andeven some of the lower cirques. If cirque development had spreadtolower areas as the ice sheet built up, cirques would be expected in peripheral areas such as the northeast and southwest, away fromthe main ice-sheet sources.This is not seen andthere is no clear evidence of time-transgressive cirque glaciation. Cirques relate to phases of glaciationwith an ELA a littlebelow that intheYounger Dryas. INTRODUCTION An attempt is made to explain the pattern of cirque glaci- ation inWales and determine if it can be related to one set of climatic conditions with consistent climatic gradients. Most attempts to understand spatial distributions of glacial cirques in a given region start fromtheworkingassumption thattheirdevelopmentwassimultaneous,relatedtoglaciers generated by a single climatic regime Porter, 1964). But does the spatial variation of the snowline make sense in relationto one climate, probably repeated, or are successive climates with different patterns of glaciation required? If the hypothesis of simultaneous cirque development is re- jected, atime-transgressive series of cirque glaciations with different snowlines is needed. This is clearly necessary for extensive areas such as northwestern North America or WesternEurope.Walesisasmallenoughareaforsimultane- ity to be possible, but varied enough that a time-transgres- sive hypothesis should alsobe considered. An understanding of the distribution of cirques as for- mer glacier sources is based on several working hypotheses about glacier mass balance in relation to topography and climateEvans,1990): 1) Glaciers form on mountains that rise above a threshold altitude; 2) The threshold altitude varies spatially with precipita- tion andtemperature regimes; 3) A lower altitude is required for glaciers on slopes with unfavourableaspects; and 4) Glacier formation is also influencedby topographic de- tails: it is less likely on mountains convex in plan and more likely in concavities, especially if they are down- wind from smooth summit areas. When glaciers occupy concavities, they often enhance them both in profile and plan, producing glacial cirques. This is believed to require warm-based ice with a surface gradient and form) sufficient to encourage rotational flow Evans,1997).Where ice builds up further, to form a moun- tain ice cap, surface gradients decline and, although some erosion may continue, the pattern of flow is unlikely to en- hance cirque concavity; over-riding ice maybevel headwall crests. It seems likely that cirques are important source areas as glaciers spread andcoalesce, but high plateaus and valley heads with or without cirques may also contribute to ice-sheet growth. Drastic changes in the spatial pattern of climate would be needed if cirque glaciers were not to grow while ice cover expanded greatly elsewhere. What is expectedunder the twoalternative hypotheses? Simultaneous cirque development implies that the spatial pattern of cirques should reflect climatic trends; cirque floors should be higher in areas that are drier or warmer and lower in those that are wetter or colder. Also, this pat- tern might resemble that of glaciers reconstructed for a given period intime. The simplest hypothesis of time-transgressive cirque glaciation is: as higher mountains become covered by ice caps, and cirque development slows down because of re- duced ice-surface gradients, previously unaffected lower mountains may undergo cirque glaciation.Thus,mountains of medium altitude could develop cirques if sufficiently isol- ated, but not if close to higher or snowier mountains. We would then expect the distribution of cirques to be spatially broader and less sensitive to altitude. Azimuthal distribu- tions would be more complicated and the lower cirques would be in drier, warmer or more peripheral locations, away from areas of ice-capbuildup. These ideas, arising from effortsto define glacial cirques and measure their form in a consistent way, are considered here in relationtoWales. A new dataset hasbeen created for 228landformsconsideredtobeglacialcirquesofvariousde- grees of development or certainty. Attributes are recorded for cirque form and for factors such as position, geology andoccupance by Late-glacial glaciers or snow beds. WALES Most of Wales consists of upland, with plateaus at various altitudes sometimes dissected into a rolling topography. Annals of Glaciology 28 1999 # International Glaciological Society 33