High‐resolution modeling of coastal freshwater discharge and glacier mass balance in the Gulf of Alaska watershed

Abstract

A comprehensive study of the Gulf of Alaska (GOA) drainage basin was carried out to improve understanding of the coastal freshwater discharge (FWD) and glacier volume loss (GVL). Hydrologic processes during the period 1980-2014 were modeled using a suite of physically based, spatially distributed weather, energy-balance snow/ice melt, soil water balance, and runoff routing models at a high-resolution (1 km horizontal grid; daily time step). Meteorological forcing was provided by the North American Regional Reanalysis (NARR), Modern Era Retrospective Analysis for Research and Applications (MERRA), and Climate Forecast System Reanalysis (CFSR) data sets. Streamflow and glacier mass balance modeled using MERRA and CFSR compared well with observations in four watersheds used for calibration in the study domain. However, only CFSR produced regional seasonal and long-term trends in water balance that compared favorably with independent Gravity Recovery and Climate Experiment (GRACE) and airborne altimetry data. Mean annual runoff using CFSR was 760 km 3 yr 21 , 8% of which was derived from the long-term removal of stored water from glaciers (glacier volume loss). The annual runoff from CFSR was partitioned into 63% snowmelt, 17% glacier ice melt, and 20% rainfall. Glacier runoff, taken as the sum of rainfall, snow, and ice melt occurring each season on glacier surfaces, was 38% of the total seasonal runoff, with the remaining runoff sourced from nonglacier surfaces. Our simulations suggests that existing GRACE solutions, previously reported to represent glacier mass balance alone, are actually measuring the full water budget of land and ice surfaces.

How much cryosphere model complexity is just right? Exploration using the conceptual cryosphere hydrology framework

“…7) for the Gulf of Alaska and finds that these temperature thresholds perform better than other sets of threshold temperatures tested. Of note, Beamer et al (2016) also find that the performance of SnowModel (the model they implement; Liston and Elder, 2006a) is insensitive to modest perturbations in the precipitation partitioning parameters. A likely reason that the model is insensitive to these parameters for this region is that Alaska has strong seasonal variations in temperature and winter temperatures are typically much colder than the mixed-phase precipitation zone.…”

“…The threshold temperatures at which all precipitation falls as snow and rain can be optimized or set. Throughout this work we set T p,s = 0 • C and T p,r = 2 • C. We choose these threshold air temperatures because Beamer et al (2016) optimizes this precipitation partitioning function (Eq. 7) for the Gulf of Alaska and finds that these temperature thresholds perform better than other sets of threshold temperatures tested.…”

“…Streamflow is then routed downhill along the flow path determined by the flow direction grid using either the "lumped" or Muskingum methods (Bedient et al, 2012). The lumped method does not allow dispersion of streamflow, whereas the Muskingum method does allow dispersion according to the formulation presented in Bedient et al (2012).…”

“…Streamflow is then routed downhill along the flow path determined by the flow direction grid using either the "lumped" or Muskingum methods (Bedient et al, 2012). The lumped method does not allow dispersion of streamflow, whereas the Muskingum method does allow dispersion according to the formulation presented in Bedient et al (2012). The results of both flow routing methods are comparable for simulating small watersheds at large time steps; however, dispersion becomes more important as the size of the watershed increases relative to the time step (Bedient et al, 2012).…”

High-resolution ice sheet surface mass-balance and spatiotemporal runoff simulations: Kangerlussuaq, west Greenland

“…SnowModel has been developed during the past three decades and tested step-wise with success on snow-and ice-covered land and ice applications: these include the Antarctic ice sheet, the GrIS, and mountain glaciers in the northern hemisphere (specifically Greenland) and southern hemisphere (e.g., Beamer et al 2016;Reiners 2002, 2006;Liston and Hiemstra 2011;Suzuki et al 2011;Suzuki, Liston, and Kodama 2015).…”