Comments on Annual Rates of Accumulation in West Antarctica

Abstract

Several maps have recently been prepared showing annual accumulation over much or all of Antarctica. These maps, based on surface measurements and on interpretations of pit stratigraphy, differ in detail but are reasonably comparable in broad aspects. Data of this type provide the basis for various recent estimates of mean annual precipitation in Antarctica ranging from 7 to 20 cm of water, with a median value near 14 and an average of about 11.5 cm. For many reasons, it is important that this figure be as accurate as possible. Well over five hundred analyses have been made of oxygen-isotope ratios (O^(18)/O^(16)) in snow samples from pits and bore holes at West Antarctic stations. These ratios display variations similar to those believed to be of seasonal origin in other areas. Greenland for example. If the variations in the accumulated Antarctic snow represent seasonal influences, they suggest annual accumulation rates 20 to 100 per cent higher than determined by other methods. The following comparative values (O^(18)/O^(16)) (determinations in parentheses) illustrate this point: South Pole, 7 (15); Byrd station, 18 (36); Little America V, 15-24 (30); Wilkes satelite, 13 (15). It is possible that annual accumulation rates in West Antarctica have generally been underestimated? It has not yet been established that the (O^(18)/O^(16) variations in the accumulated snow and firn or Antarctica are definitely of seasonal origin; some other type of cycle or influence may be represented. However, the consistency and large magnitude of those variations show that they are not the product of mere chance; a basic control of some sort is reflected. In our present state of knowledge seasonal control seems the most likely. The short time interval over which surface measurements of accumulation have been made and the uncertainties attending interpretations of pit stratigraphy in Antarctica arc such that every opportunity should be taken to check these methods by independent means, such as (O^(18)/O^(16) variations.