What do accumulation records of single ice cores in Greenland represent?

Abstract

It is investigated to what extent multiannual accumulation time series from Greenland reflect dynamic and thermodynamic processes and how representative single accumulation series are for the entire ice sheet. Furthermore, it is examined whether accumulation is related to low‐level atmospheric temperatures. For this purpose, two kinds of regression models are developed which linearly relate multiannual accumulation records to meteorological mean fields. Seven ice cores from north to central Greenland and the NCEP Reanalysis data are used for the period from 1948 to 1992. In order to reduce noise, the data are smoothed with a weighted 5‐year running mean. The downscaling technique is based on a stepwise multiple linear regression. One group of regression models distinguishes between dynamic and thermodynamic atmospheric effects. For that reason, stream functions are used in a first step to describe the dynamically controlled accumulation, whereas the thermodynamically controlled accumulation is determined by temperature in a second step. For six of the ice cores, these regression models describe more than 56% of the variability of the smoothed accumulation series, confirming that they represent to a large extent atmospheric states. Multiannual accumulation variability is found to dominantly represent circulation variability. However, the circulation fields that are linked with accumulation show marked differences among the cores concerning the represented seasons, areas, and structures. Thus local accumulation generally represents only regional‐scale climate features, which are probably to a great extent influenced by orography. Furthermore, regression models using only 700 hPa temperature as predictor show that a general linear relationship between accumulation and temperature does not exist over this 45‐year time interval. Therefore paleoaccumulation rates derived from isotopic temperatures should be interpreted with caution. Moreover, it is not reasonable to describe accumulation by means of temperature in mass balance models for the Greenland ice sheet in decadal timescales.