An 11-year record of mass balance of Brewster Glacier, New Zealand, determined using a geostatistical approach

Nicolas J Cullen,Ruzica Dadic,Andrew Lorrey,Dorothea Stumm,Brian Anderson,Pascal Sirguey,Sean J Fitzsimons,Huw J Horgan,Jonathan P Conway,Andrew Mackintosh

doi:10.1017/jog.2016.128

Abstract

ABSTRACTRecognising the scarcity of glacier mass-balance data in the Southern Hemisphere, a mass-balance measurement programme was started at Brewster Glacier in the Southern Alps of New Zealand in 2004. Evolution of the measurement regime over the 11 years of data recorded means there are differences in the spatial density of data obtained. To ensure the temporal integrity of the dataset a new geostatistical approach is developed to calculate mass balance. Spatial co-variance between elevation and snow depth allows a digital elevation model to be used in a co-kriging approach to develop a snow depth index (SDI). By capturing the observed spatial variability in snow depth, the SDI is a more reliable predictor than elevation and is used to adjust each year of measurements consistently despite variability in sampling spatial density. The SDI also resolves the spatial structure of summer balance better than elevation. Co-kriging is used again to spatially interpolate a derived mean summer balance index using SDI as a co-variate, which yields a spatial predictor for summer balance. The average glacier-wide surface winter, summer and annual balances over the period 2005–15 are 2484, −2586 and −102 mm w.e., respectively, with changes in summer balance explaining most of the variability in annual balance.

Highlights

Monitoring surface mass balance of mountain glaciers is important as fluctuations in glacier mass over time have hydrological implications for water resource users and are used to infer changes in climate
Cross validation of the co-kriging of snow depth index (SDI) and mean summer balance index (MSBI) results in mean errors of 0.11 mm and 4.5 mm w.e., respectively, which are indicative of negligibly biased models
The determination of the co-kriging standard error is rigorous in a geostatistical sense, the quantification of the uncertainty in the upper accumulation area needs to be further scrutinised

Summary

Introduction

Monitoring surface mass balance of mountain glaciers is important as fluctuations in glacier mass over time have hydrological implications for water resource users and are used to infer changes in climate. A globally coordinated effort to obtain mass-balance measurements has enabled glacier-wide results from >250 glaciers to be documented (Zemp and others, 2015), with >4000 annual observations available spanning the past seven decades. To address the issue of Southern Hemisphere glaciers being heavily under-represented in the global glacier monitoring network, a mass-balance programme on Brewster Glacier in the Southern Alps of New Zealand was initiated in 2004. The Southern Alps are surrounded by vast areas of ocean and are strongly influenced by both subtropical and polar air masses, with the interaction of these contrasting air masses in the prevailing westerly airflow resulting in synoptic scale atmospheric circulation having a strong influence on glacier behaviour (Fitzharris and others, 1992, 1997; Clare and others, 2002; Gillett and Cullen, 2011). The massbalance programme on Brewster Glacier has provided a platform from which to assess glacier sensitivity to atmospheric forcing in the Southern Alps (Anderson and others, 2010; Gillett and Cullen, 2011; Conway and Cullen, 2013, 2016; Conway and others, 2015; Cullen and Conway, 2015)

Objectives

Methods

Results

Conclusion