Open Collections

Winter-Billington, Alexandra

University of British Columbia

Meltwater from glaciers contributes to hydro-electricity and agricultural production in some places, and model predictions of glacier melt can inform water management decisions. However, predictions are inherently uncertain. Accurately constrained uncertainty can be key to success of management decisions. Few studies have estimated the uncertainty in predictions of debris-covered glacier (DCG) melt. This study evaluated the accuracy of predictions made using both empirical and physically based sub-debris melt models. Empirical mixed-effects relations were fit using data from multiple DCG. The most accurate model in cross-validation, KP1, was fit using melt factors calculated with on-site and MERRA-2 air temperature. KP1 predicts melt factors on the basis of an exponential relation with debris thickness (fixed effect), and site and year of observation (random effects). Model KP1 can be used to estimate melt at any site with a known margin of error; however, considerable uncertainty remained, estimated at ± 98 % of the predicted value. Prediction uncertainty was investigated using a new physically based model, SHAW-Glacier. Based on the Simultaneous Heat and Water (SHAW) transport model (Flerchinger, 1987), SHAW-Glacier is the most physically complete representation of DCG melt yet reported. Predictions made using SHAW-Glacier aligned with observations in general, but could not account for observed ablation at North Changri Nup glacier. Analyses revealed a non-linear sensitivity in modelled ablation to precipitation phase, suggesting a sensitivity to snow cover. Further analyses supported the hypotheses that sub-debris melting is sensitive to summer snow cover distribution and rainfall. Uncertainty in precipitation phase and snow distribution outweighed the effect of debris thickness, and outweighed the effect of all the other physical parameter values, combined. This thesis shows that interactions between debris thickness and meteorological conditions are key to understanding the spatiotemporal variation of sub-debris melt. A fixed relation between sub-debris melt and debris thickness may not hold under varying meteorological conditions or snow distribution patterns. Moreover, patchy snow cover is a source of considerable prediction uncertainty that should be included in glacier-scale and regional model studies. SHAW-Glacier is a flexible tool that may be used to improve understanding of the complex interactions between variables at different sites.

Text

2023-10-31

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/86416

Geography

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/