SOME OBSERVATIONS ON NEEDLE ICE By D. M. LAWLER Department of Geography, University of Birmingham T HE recent compilation by Woodley (1988) raises the interesting topic of needle ice (pipkrake), which is becoming increasingly recognized as a significant process in many areas of environmental science. I am writing to point out how common needle ice actually is and to help bring to the attention of meteorologists the quite considerable body of scientific work on the phenomenon now available and which may not be generally familiar. The opportunity is also taken to call for further observations of needle ice, to correct some oft-repeated inaccuracies, and to comment on the slight terminological confusion which exists. Outcalt (1971 p.394) defines needle ice as "filaments of ice about 1rnm? in cross section and up to 8-lOcm in length, (which) are formed by the segregation of ice near the ground surface during calm, clear, winter evenings, when the near-surface soil has been either unfrozen or thawed during the hours of daylight". The ice filaments usually grow orthogonally to the ground surface, and lengths up to 40cm have been recorded [Krumme 1935 (cited in Troll 1958)j. Although Petkovsek's photograph (communicated by Woodley (1988» shows it growing from the branch of a tree, needle ice is, perhaps, more usually found issuing from moist, bare, soil following a night frost (e.g. Fig. I). Thanks largely to the detailed work of Outcalt (1970; 1971; et seq) and Soons and Greenland (1970) we now know that the surface of an appropriately-textured, 'frost- susceptible', moist soil has only to cool to around - 2°C for needle ice to form, as long as a supply of unfrozen moisture is freely available to migrate to the freezing front. Too fine a soil, however, will inhibit the movement of moisture to 'feed' the base of the ice needles, while soils which are too coarsely-grained will not be able to establish the crucial suction potential required to 'pull' moisture to the freezing front (Washburn 1979). These growth conditions, in practice, are not too difficult to fulfil, and needle ice is therefore widespread throughout the world (Troll 1958), with all continents represented. It is much more common than might be supposed. From a recent compilation of over 100 global sightings, however, Lawler (l988a) has shown that reasonably well-defined altitudinal and latitudinal limits seem to exist, and these describe an inverted V-shaped envelope (Fig. 2). Note that the British Isles (rv 50° -59°N), as well as Petkovsek's Yugoslavian observations ( rv46°N; 600m ASL), sit firmly within what might be called the 'zone of needle-ice potential' (Fig. 2). Work is continuing to relate this envelope distribution to climatic indices. Further observations are required, though, to define more clearly the zonal boundaries, and to strengthen the distribution in certain areas (e.g. the high and low latitudes and the southern hemisphere), and I should be very grateful for notification of any sightings made by readers. The details required include altitude, latitude, and longitude of the site, together with, if possible, date of observation, needle-ice characteristics (e.g. length, presence of included sediment), and some general comments on the site (e.g. soil characteristics, aspect). A proforma can be sent on request. Different scientists have become interested in needle ice for different reasons: micrometeorologists (e. g. Outcalt 1971) because it provides a useful vehicle with which to study the delicate balance and flux of energy and moisture; geomorphologists because of its erosional impact on hillslopes (e. g. Soons and Rayner 1968) and river banks (e.g. Gardiner 1983; Lawler 1987) and its formative role in a whole suite of periglacial landforms such as soil stripes (e.!? Heilbronn and Walton 1984); agronomists and ecologists due to its destructive effects on young seedlings (e.g. Brink 1964); and engineers, following recognition of attendant damage to road surfaces, road-cuts, and structures te.g, Hursch 1949). Within geomorphology, especially, needle ice has, since the time of Le Conte (1850), been recognized as a potentially significant process in 406