landscape and climate W. M. Davis based his classic geographical cycle or cycle of erosion on the certainty that ‘the greater part of the land surface has been carved by … the familiar processes of rain and rivers, of weather and water’. He defined a ‘normal’ climatic region as ‘not so dry but that all parts of the surface have continuous drainage to the sea, nor so cold but that the snow of winter all disappears in summer’. His definition betrays a certain provinciality that was common in eastern North America and western Europe in the late nineteenth and early twentieth centuries. In the Davisian scheme, aridity and glaciation were viewed as ‘climatic accidents’. The concept of ‘normal’ and ‘accidental’ landscape evolution was perpetuated in the titles of two widely read books by the distinguished New Zealand geomorphologist Sir Charles Cotton:
Landscape as developed by the processes of normal erosion (1941, 1948) and
Climatic accidents in landscape-making (1942, 1947), as well as by the Belgian textbook by Paul Macar:
Principes de géomorphologie normale: étude des formes du terrain des régions à climat humide (1946). Although Cotton quoted Davis to the effect that climatic accidents include changes from humid to arid and from cooler to warmer conditions, in his book on climatic accidents he considerably broadened these simple categories to include dry and dry-seasonal climatic landscape types in addition to glaciated landscapes.
Climatic morphogenesis
In the decades that followed the Second World War, the concept of climatic morphogenesis evolved rapidly, especially in western Europe, beginning with Pierre Birot's brief book
Le cycle d'érosion sous les différents climats (1960; English translation 1968) and followed by J. Tricart's
Introduction à la géomorphologie climatique (1965; English translation 1972) and J. Büdel's
Klima-geomorphologie (1977; English translation 1982). On a present-day geomorphologist's bookshelves will be a dozen or more monographs dealing with the landscapes of specific climates, among them the glacial, periglacial, semi-arid savannah, humid tropical, and highland regions.
A morphogenetic region has been defined as being characterized by certain climatic factors that produce a distinct combination of geomorphological processes, which in turn may produce a landscape distinct in appearance from other regions. The concept of morphogenetic regions dominated by processes subordinates the roles of structure and time in landscape evolution. Followed to the extreme conclusion, it implies that each climatic region produces a unique assemblage of landforms independent of time and structure. However, both Cotton and Birot followed the basic Davisian concept of a cycle of erosion, deducing climate-determined variants of the ‘normal’ cycle. Cotton offered deduced characteristic profiles of mature landscapes in arid, semi-arid, savannah, and hot-humid cycles of erosion. Birot, as the title of his book clearly states, offered alternative schemes of cyclic landscape evolution under a normal climate, a tropical climate, arid and semi-arid climates, a climate of alternating wet and dry seasons or of alternating wet and dry climates, and a periglacial climate. Few, if any, geomorphologists would accept an extreme view of climatic morphogenesis independent of structure and time, although all would agree that even an untrained tourist will recognize the distinctive landscapes of periglacial, desert, or tropical rainforest regions, among others.
Using the five traditional Köppen climatic divisions, the non-ice-covered land surface is divided according to the following approximately equal percentages: tropical humid, 20 per cent; arid and semi-arid, 26 per cent; (warm) temperate humid, 16 per cent; boreal forest with snow, 21 per cent; and cold, treeless snow regions, 17 per cent. By other compilations, about 10 per cent of the land area is now ice-covered, 22 to 25 per cent is underlain by perennially frozen ground, and up to 33 per cent is characterized by drainage that does not reach the sea, a useful definition of aridity. Unique combinations of geomorphological processes are imposed on each of these and other landscape categories. Current debate concerns the extent to which they are unique morphogenetic regions.
Late Cenozoic climate change
The new geomorphological paradigm that has grown out of studies of Late Cenozoic climates is that the magnitude and frequency of climate changes during the past 25 million years, and certainly during the past 2.7 million years, have greatly exceeded the rate of landscape responses. Thus, all but the youngest landscapes are palimpsests, written over by a variety of successive or alternating sets of climate-related processes. Up to 30 per cent of the land area has been glaciated repeatedly during the past 1.7 million years. Perhaps 40 per cent of the land area has been subjected to periglacial conditions recently enough for relict landforms to persist. Subtropical deserts and tropical savannahs and rainforests have similarly expanded and contracted, imposing their morphogenetic overprint on older landscapes.
Julius Büdel proposed that fully 95 per cent of the present European landscape is relict, dominated to progressively greater degrees by glaciation (his subglacial relief zone), periglacial processes (his polar zone of extreme valley cutting), and savannah processes (his subtropical zone of extreme planation). He argued that the landscape of all continents, extending from the Equator into the polar regions, is dominated by relict Palaeogene or even Cretaceous savannah etchplains and pediplains. As latitudinal climate zonation became more pronounced during the Neogene (Fig. 1), the zone of etchplains and inselbergs contracted into the tropics, but left behind vast upland surfaces of low relief that have subsequently been modified by periglacial valley incision at least as far south as 40°N, and to a lesser degree by glacial erosion. The relict low-relief etchplains, some of which formed over an interval of 100 million years, are extremely durable compared to the much younger Pleistocene periglacial and glacial landscapes. Although Büdel's hypotheses of relict climate-controlled landforms represent only one end member in a broad spectrum of opinion about climate morphogenesis, they provide a useful set of testable propositions for continuing geomorphological analysis and debate.
No-analogue climates
When climate zones migrate poleward or equatorward, they cannot retain their entire character. The tundra zone of today, broadly coincident with permafrost, is to be found primarily in high latitudes, where seasonal changes in length of daylight and darkness are extreme. Biological and geomorphological activity are strongly dominated by seasonal, rather than diurnal, photoperiodicity. The annual production of an active solifluction layer above permafrost, and the annual spring breakout of Arctic rivers, give that landscape its strong climatic imprint. But what was the periglacial landscape like in Illinois, USA, or on the Korean peninsula at 35–40 °N latitude, peripheral to full-glacial ice sheets? Relict permafrost soil structures demonstrate average annual soil temperature of –5°C or less. But with the summer sun only 15° from overhead, and even the winter sun well above the horizon, how did mid-latitude periglacial morphogenesis work? Were there seasonal ‘January thaws’? How cold did the winters have to be to counteract the inevitably warmer summers? The modern high-latitude tundra is not an appropriate analogue. In fact, none exists. Latitude is a nearly invariant climatic parameter, and so therefore are sun angle and length of day. High-altitude, low-latitude regions of permafrost equally offer poor climatic analogues to former mid-latitude periglacial regions.
An even more extreme case of no-analogue climate is the Palaeogene ‘greenhouse’ climate that extended to the shores of an ice-free Arctic Ocean. Canadian and Danish palaeobotanists describe fossil trees that grew within 10–20° of the North Pole, yet lack any indication of frost-damaged tree rings. Where is the modern analogue for a high-latitude coastal plain with rivers flowing throughout the year, abundant rainfall, and an open spruce–birch woodland, yet with 6 months of near-total darkness? Again, none exists, yet in such a climate the crystalline shields of Canada and Scandinavia developed a deep saprolitic weathered surface layer that was subsequently removed by glacial erosion.
We must ask hard questions of those who hypothesize relict climatic morphogenesis in latitudes well beyond those in which the characteristic morphogenetic processes are now active. Is the modern tropical savannah the correct analogue for Palaeogene Europe? Could seasonal rainfall alternating with extended drought create etchplains at 40–55 °N latitude? Did the periglacial landscape of southern France look like the modern arctic tundra, with similar morphogenesis? These questions open a fertile field for future research on landscape and climate.
Arthur L. Bloom
Bibliography
Bloom, A. L. (1998) Geomorphology: a systematic analysis of Late Cenozoic landforms (3rd edn), Chapters 4 and 18. Prentice Hall, Englewood Cliffs, New Jersey.
Büdel, J. (1982). Climatic geomorphology (Trans. L. Fischer and and D. Busche ). Princeton University Press.
