The Hydrological System and Climate of Brewster Glacier, Tititea Mt Aspiring National Park, Southern Alps, Aotearoa New Zealand, in the Context of Climate Change.
Temporal and spatial variability of stream discharge is directly related to variation in local climate, and this in turn is related to both regional and global atmospheric circulation and climate change. The relationship is complicated in glacierised catchments. This study aims to identify relationships between discharge from Brewster Glacier proglacial stream and both local atmospheric variables and national atmospheric circulation patterns. An attempt is made to quantify these relationships using statistical models and tests in order that prediction of discharge with climate change could be made using local weather forecasts and national circulation indices. The nature of the subglacial drainage system is also investigated with particular focus on its structural evolution from summer to autumn. It is found that shortwave radiation, wind speed and relative humidity are consistently the most important variables in prediction of discharge and that wind speed is most important during summer while air temperature is most important in autumn. It is concluded that the importance of precipitation is greater than indicated by the results which were influenced by covariance in the records. A multiple regression model for summer discharge predicts up to 85% of variation in the proglacial stream hydrograph and for autumn 60%. Low overall energy inputs during autumn result in lesser sensitivity of discharge to variation in environmental conditions. It is concluded that the subglacial drainage system is highly arborescent over both summer and autumn and that little, if any, evolution occurs through these seasons. A qualitative relationship is established between discharge production at Brewster Glacier proglacial stream and national atmospheric circulation indices; highest average discharge occurs during northwesterly cyclonic conditions, when the turbulent heat fluxes and precipitation dominate discharge production, and lowest during southeasterly anticyclones when total energy inputs are low. The multiple regression models are used to estimate changes in discharge over the next 20 years given predicted changes in air temperature and precipitation, and it is found that the models lack the sensitivity required for accurate predictions.