Temperature-index modeling of mass balance and runoff in the Valdez glacier catchment in 2012 and 2013

Abstract

Glaciers play an important role in both storage and generation of runoff within individual watersheds. The Valdez Glacier catchment (342 km²), located in southern Alaska in the Chugach mountains off of Prince William Sound, is characterized by large annual volumes of rain- and snowfall. As Valdez Glacier and other glaciers within the catchment (comprising 58% of the catchment area) continue to melt in a warming climate, it is unclear how the runoff will be affected. Temperature-index modeling is one method used to estimate glacier mass balance and runoff in highly glacierized catchments, and may be suitable for predicting future runoff regimes. In this study, we used a combination of field measurements (air temperature, glacier mass balance, streamflow, and ground-penetrating radar (GPR)-derived snow water equivalent (SWE) from a parallel study) and modeled climate data (PRISM) to a) calibrate a temperature-index model to glacier mass balance in 2012; b) validate the model to laser altimetry; and c) calibrate a temperature-index model to runoff measurements in fall of 2012 and in spring, summer and fall of 2013. We calibrated the snow-radiation coefficient (r_snow), ice-radiation coefficient (r_ice), and melt factor (MF) of the temperature-index model to glacier mass balance measurements from 2012. Using the calibrated- r_snow, r_ice, and MF (i.e. r_snow, r_ice, and MF = 0.20, 0.50 and 4.0, respectively), we calculated 2012 annual glacier mass balance (Ba) at 0.05 ± 0.49 meters water equivalent (m w.eq.). We next validated the model to 2012 laser altimetry annual glacier mass balance estimates (Ba = 0.20 ± 0.6 m w.eq.). We then modeled glacier mass balance in 2013 using r_snow, r_ice, and MF from the 2012 calibration. The model underestimated summer glacier mass balance in 2013, resulting in annual glacier mass balance (Ba = 0.55 m w.eq.) that did not fall within the 2013 laser altimetry annual balance estimate (Ba = -1.15 +0.29/-0.30 m w.eq.). We therefore re-calibrated MF to 2013 laser altimetry measurements, resulting in an annual glacier mass balance (Ba) of -1.10 ± 0.49 m w. eq. We next calibrated the storage constants of the runoff model to hydrographs from mid-September until mid-October 2012, and from May until October 2013, with r_snow, r_ice, and MF set to values from the 2012 glacier mass balance calibration. Total modeled runoff in mid- September until mid-October 2012 was within 3% of measured runoff (E- and lnE- were 0.54 and 0.76, respectively). Modeled runoff in 2013 was calculated to within 5% of 2013 runoff measurements (E- and lnE-values of 0.79 and 0.70, respectively). We next modeled runoff in 2013 using MF from the 2013 glacier mass balance calibration to laser altimetry (i.e. MF = 7.0). The fit of 2013 modeled to measured runoff was reduced (E- and lnE- values of 0.44 and 0.54, respectively), suggesting that additional glacier mass balance measurements are necessary in 2013 in order to properly calibrate the model. Results indicate that glacier melt parameters likely vary inter-annually. Therefore, the temperature-index model is capable of modeling both glacier melt and runoff in a maritime catchment, provided that ablation stake, air temperature, precipitation, and streamflow measurements are available for the simulation period.