Quantifying parameter uncertainty in a large-scale glacier evolution model using Bayesian inference: application to High Mountain Asia

David R Rounce,Margaret B Short,David E Shean,Douglas J Brinkerhoff,Regine Hock,Tushar Khurana

doi:10.1017/jog.2019.91

David R Rounce, Margaret B Short + Show 4 more

Open Access

https://doi.org/10.1017/jog.2019.91

Copy DOI

Abstract

AbstractThe response of glaciers to climate change has major implications for sea-level change and water resources around the globe. Large-scale glacier evolution models are used to project glacier runoff and mass loss, but are constrained by limited observations, which result in models being over-parameterized. Recent systematic geodetic mass-balance observations provide an opportunity to improve the calibration of glacier evolution models. In this study, we develop a calibration scheme for a glacier evolution model using a Bayesian inverse model and geodetic mass-balance observations, which enable us to quantify model parameter uncertainty. The Bayesian model is applied to each glacier in High Mountain Asia using Markov chain Monte Carlo methods. After 10,000 steps, the chains generate a sufficient number of independent samples to estimate the properties of the model parameters from the joint posterior distribution. Their spatial distribution shows a clear orographic effect indicating the resolution of climate data is too coarse to resolve temperature and precipitation at high altitudes. Given the glacier evolution model is over-parameterized, particular attention is given to identifiability and the need for future work to integrate additional observations in order to better constrain the plausible sets of model parameters.

Highlights

Glacier mass loss in the 100 years has critical implications for sea-level change and water resources across the globe
Future projections of glacier mass loss for a given General Circulation Models (GCMs) and Representative Concentration Pathways (RCPs) vary considerably depending on the model, which is attributed to differences in the model physics, calibration data/methods, and input data (Hock and others, 2019)
Three chains of 25,000 steps for 1,000 glaciers in each of the three major RGI regions in High Mountain Asia were used to evaluate the performance of the Markov chain Monte Carlo (MCMC) method and determine the acceptable number of steps that should be used when calibrating all the glaciers

Summary

Introduction

Glacier mass loss in the 100 years has critical implications for sea-level change and water resources across the globe. Future projections of mass loss from several glacier evolution models using output from an ensemble of General Circulation Models (GCMs) and Representative Concentration Pathways (RCPs) estimate the increase in sea level due to glaciers by 2100 relative to 2015 to range from 94 ± 25 mm SLE (RCP2.6) to 200 ± 44 mm SLE (RCP8.5) (Hock and others, 2019). While the rate of sea-level change is dominated by regions with the most glacier mass (Alaska and the arctic regions), glacier mass loss in other regions may fundamentally alter the quantity and timing of glacier runoff, thereby affecting water resources (Bliss and others, 2014; Huss and Hock, 2018). Excluding mass-balance sensitivity models (Slangen and others, 2012), these sparse observations are used to calibrate between two and seven parameters in each model, which affect the air temperature, precipitation, and mass balance (Table 1). Each model has at least one parameter affecting the mass balance (e.g., degree-day factors, temperature sensitivity, or mass-balance gradient), and seven of the nine models cited above use a precipitation correction factor to correct for potential biases in the climate data

Results

Discussion

Conclusion