 |
| Titel |
Linking surface energy balance calculations and Temperature Index models of surface melt: Revision of the Positive Degree-Day (PDD) methodology for the Greenland Ice Sheet |
| VerfasserIn |
L. M. Wake, S. J. Marshall, B. Lecavalier, G. A. Milne, P. Huybrechts, M. J. R. Simpson, N. Bayou |
| Konferenz |
EGU General Assembly 2012
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 14 (2012) |
| Datensatznummer |
250064336
|
|
|
|
|
|
| Zusammenfassung |
| Positive Degree-Day (PDD) methodology (Braithwaite and Olesen, 1989; Reeh, 1991) is
widely used in conjunction with observationally-derived Degree-Day Factors (DDFs) for
snow and ice in order to simulate ice-sheet wide ablation rates using mean monthly
temperature as the only input. Monthly PDD totals are calculated using the assumption that
the monthly temperature distribution follows a Gaussian relationship with a spatially and
temporally invariable standard deviation (Ïăm), typically in the range of 4-5oC. DDFs for
snow and ice used in ice sheet modelling are usually fixed at ~3 and 8 mm w.e. oC-1
day-1 respectively, but field observations show that these can vary by at least a
factor of two depending on the albedo characteristics of the glacier surface (Hock,
2003).
At odds with the assumption of constant Ïăm, it has been shown that temperature
variability is reduced at temperatures close to or above the melting point, due to thermal
(latent heat) buffering and the maximum temperature of 0oC for a melting snow/ice surface
(e.g., Marshall and Sharp, 2009). Analysis of hourly temperature data from 22 GC-Net
stations (Steffen and Box, 2001) spanning the period 1995-2010 shows that observed Ïăm
follows a quadratic relationship with observed average monthly temperature. Comparisons of
calculated and observed monthly PDD totals from GC-Net locations show that
current assumptions of Ïăm = 4-5oC can overestimate monthly PDD totals by 25%
on average, compared to ~3% for the new methodology using a spatially varying
Ïăm.
In the absence of extensive field measurements, ‘theoretical’ daily melt rates are
calculated at several GC-Net locations using available field data to estimate the components
of the daily surface energy budget (Net radiation, sensible and latent heat and subsurface
energy flux). Subsequently, ‘theoretical’ DDFs are evaluated as a function of surface albedo.
Our results suggest that future studies should consider DDFs for snow and ice of 3-7 mm w.e.
oC-1 day-1 and > 9 mm w.e. oC-1 day-1 respectively, alongside the use of a spatially and
temporally variable Ïăm, which can be parameterized as a function of monthly mean
temperature.
We demonstrate the impact of our new methodology and previous assumptions of
fixed Ïăm on millennial-scale changes in Greenland Ice Sheet volume and resulting
global and local RSL predictions using a 3D thermomechanical ice model where a
spatially and temporally variable Ïăm acts to reduce the magnitude and variability
associated with surface melt. These results demonstrate a need for additional empirical
derivations of Degree-Day Factors in Greenland and their relationship with surface
albedo in order to compliment the purely theoretical approach taken in this study. |
|
|
|
|
|
|