TY - JOUR
T1 - Statistical EOF analysis of spatiotemporal glacier mass-balance variability
T2 - a case study of Mittivakkat Gletscher, SE Greenland
AU - H. Mernilda, Sebastian
AU - Beckerman, Andrew P.
AU - Knudsen, Niels Tvis
AU - Hasholt, Bent
AU - Yde, Jacob C.
PY - 2018/1/2
Y1 - 2018/1/2
N2 - An Empirical Orthogonal Function (EOF) variance analysis was performed to map in detail the spatiotemporal variability in individual stake mass-balances (ba) on Mittivakkat Gletscher (MG)–in a region where at present five out of ~20.000 glaciers have mass-balance observations. The EOF analysis suggested that observed ba was summarized by two modes: EOF1 and EOF2 represented 80% (significant) and 6% (insignificant) of the explained variance, respectively. EOF1 captured a decline in ba that was uniformly distributed in space at all stakes. The decline was correlated with albedo observations and air temperature observations from nearby stations. EOF2, however, described variations in ba that were heterogeneously distributed among stakes and associated with local slope and aspect. Low elevation stakes (~<400 m a.s.l.) showed relatively negative (out of phase) correlation and higher elevated stakes relatively positive (in phase) eigenvector correlation values with EOF2. Such relatively negative and positive eigenvector correlation values were present where the constituted of exposed glacier ice or snow cover, respectively. The results from this study show how EOF analyses can provide information on spatiotemporal patterns of glacier mass-balance. Understanding such detailed variabilities in mass-balance on a Greenlandic glacier is of interest because a fifth of the Arctic contribution from glaciers and ice caps to sea-level rise originates from Greenland.
AB - An Empirical Orthogonal Function (EOF) variance analysis was performed to map in detail the spatiotemporal variability in individual stake mass-balances (ba) on Mittivakkat Gletscher (MG)–in a region where at present five out of ~20.000 glaciers have mass-balance observations. The EOF analysis suggested that observed ba was summarized by two modes: EOF1 and EOF2 represented 80% (significant) and 6% (insignificant) of the explained variance, respectively. EOF1 captured a decline in ba that was uniformly distributed in space at all stakes. The decline was correlated with albedo observations and air temperature observations from nearby stations. EOF2, however, described variations in ba that were heterogeneously distributed among stakes and associated with local slope and aspect. Low elevation stakes (~<400 m a.s.l.) showed relatively negative (out of phase) correlation and higher elevated stakes relatively positive (in phase) eigenvector correlation values with EOF2. Such relatively negative and positive eigenvector correlation values were present where the constituted of exposed glacier ice or snow cover, respectively. The results from this study show how EOF analyses can provide information on spatiotemporal patterns of glacier mass-balance. Understanding such detailed variabilities in mass-balance on a Greenlandic glacier is of interest because a fifth of the Arctic contribution from glaciers and ice caps to sea-level rise originates from Greenland.
U2 - 10.1080/00167223.2017.1386581
DO - 10.1080/00167223.2017.1386581
M3 - Journal article
SN - 0016-7223
VL - 118
SP - 1
EP - 16
JO - Geografisk Tidsskrift
JF - Geografisk Tidsskrift
IS - 1
ER -