DOI: 10.3112/erdkunde.2009.03.02 ISSN 0014-0015 http://www.giub.uni-bonn.de/erdkunde PERTURBATION OF CLIMATIC RESPONSE AT MARITIME GLACIERS?
Stefan Winkler and atle neSje With 13 figures and 6 tables
Received 7 February 2009 ∙ Accepted 29 July 2009
Summary: �etailed analyses �� �ass�balance and length�variati�n data �r�� �ariti�e ��untain glaciers in s�uthern ��r��etailed analyses �� �ass�balance and length�variati�n data �r�� �ariti�e ��untain glaciers in s�uthern ��r- way reveal that s��e �requent assu�pti�ns �� the relati�nship between �ass�balance data and ter�inus resp�nse need t�
be rec�nsidered. In particular, the p�ssibility �� a ‘regi�e shi�t’ in the �ass�balance drivers and the virtual absence �� any
�r�ntal reacti�n ti�e due t� a perturbati�n �� the dyna�ic resp�nse �� the glacier t�ngue �ccurring ar�und c. A� 2000 are discussed. Alth�ugh ab�ve�average su��er air te�peratures una�bigu�usly caused the ��st recent retreat, it is n�t clearly linked t� �ass�balance data. Further��re, the �ariti�e glaciers �� s�uthern ��rway are in general n�t entirely deter�ined by air te�perature changes. Relative c�ntributi�ns �� the winter balance t� the annual net �ass balance variati�ns were high during the last decades �� the 20th century and a c�nsiderable increase in ice �ass during the 1990s was caused by increased winter precipitati�n. There��re, the para�eter ‘annual air te�perature’ cann�t be applied t� explain glacier length variati�ns.
Zusammenfassung: �ie detaillierte Analyse v�n Massenbilanz� und Längenänderungs��atenreihen �ariti�er H�chge- birgsgletscher in Südn�rwegen zeigt, dass häufig verwendete Annah�en über die Beziehung zwischen Massenbilanzda- ten und Längenänderungen einer kritischen Überprü�ung unterz�gen werden �üssen. Insbes�ndere die Möglichkeit eines
„Regi�ewechsels“ innerhalb der �assenbilanzdeter�inierenden Fakt�ren und die augenscheinliche Abwesenheit jeglicher
�r�ntalen Reakti�nszeit durch Störung der dyna�ischen Reakti�n der Gletscherzunge seit unge�ähr de� Jahr 2000 werden diskutiert. Obw�hl überdurchschnittliche s���erliche Lu�tte�peraturen unzwei�elha�t den aktuellen Rückzug verursacht haben, ist jener nicht deutlich an die Massenbilanzdaten gek�ppelt. Außerde� werden die �ariti�en Gletscher Südn�r- wegens generell nicht ausschließlich durch Lu�tte�peraturvariati�nen gesteuert. �er relative Beitrag der Winterbilanz an den jährlichen �ett�bilanzschwankungen während der letzten Jahrzehnte des 20. Jahrhunderts war h�ch, ein erhebliches Eis�assenwachstu� in den 1990er Jahren wurde durch angestiegene Winterniederschläge verursacht. �aher kann der Para-
�eter „Jahres�ittelte�peratur“ hier nicht angewendet werden, u� die Gletscherlängenänderungen zu erklären.
Keywords: Glacier variati�ns, cli�ate change, �ariti�e ��untain glaciers, J�stedalsbreen, western ��rway
1 Introduction
M�untain glaciers are key indicat�rs �� gl�bal cli�ate change (IPCC 2007), and ��r several aspects �� sustainable devel�p�ent in high���untain regi�ns (hydr��electric en- ergy, water supply, t�uris�, etc.) it is crucial t� esti�ate �u- ture glacier variati�ns (Huber et al. 2005). Changes in glacier v�lu�e, area, and length are deter�ined by the cli�ate and related �ass flux/glacier fl�w. There��re, the interacti�ns and relati�nships between individual �ete�r�l�gical and glaci�l�gical para�eters need t� be kn�wn be��re any ��d- el can be applied. M�st ��dels assu�e that these �act�rs are c�nstant �ver ti�e and thus apply l�ng�ter� �ean data
��r c�nstructi�n �� ��del alg�rith�s (bamber and Payne
2004). Glacier length variati�ns are assu�ed t� be a dyna�- ic resp�nse �� the glacier t�ngue t� �ass�balance changes, and si�ulated acc�rdingly (OerlemanS 2001; HOOke
2005). Anal�g�usly, cu�ulative glacier length changes are c�llected t� derive esti�ati�n ��r annual air te�perature in-
crease during the 20th century (OerlemanS 1994, 2005). In this paper, we highlight new e�pirical results �r�� �ariti�e s�uthern ��rway that �ay be c�ntradict�ry t� s��e �� the
�requently applied assu�pti�ns regarding the relati�nship between �etr�l�gical/cli�ate data and glacier resp�nse, es- pecially the representativeness �� l�ng�ter� average data, c�nstant ter�inus reacti�n ti�es, and the c�upling �� net
�ass�balance data with length changes. In particular, we want t� present a hyp�thesis that there is the p�ssibility �� a regi�e shi�t in the �ass�balance drivers and/�r the �echa- nis�s �� ter�inus reacti�n at these glaciers.
2 The study area and its glaciological regime S�uthern ��rway is a regi�n with detailed l�ng�ter� glaci�l�gical data sets (andreaSSen et al.
2005; ZemP et al. 2008). Within a distance �� �nly 180 k� al�ng an E�W pr�file, i�p�rtant spatial di�-
�erences in �ass�balance characteristics, net �ass�
balance trends, and length variati�ns take place (Figs. 1, 2). �espite di��erent abs�lute cu�ulative net �ass�balance trends �ver the past decades (Fig.
3), relative trends �� seas�nal �ass�balance data reveal s��e parallels (Tab. 1). Pr��inent p�sitive
�r negative departures �� annual seas�nal and net
�ass balances �requently �ccur al��st si�ultane-
�usly thr�ugh�ut s�uthern ��rway (Fig. 4). This pattern w�uld n�t be present with regi�nally di��er- ent cli�ate trends �r a high degree �� spatial di��er- entiati�n �� the recent cli�ate fluctuati�ns within this regi�n. There��re, the existing net �ass�bal- ance di��erences cann�t ulti�ately be related t� a regi�nally di��erent cli�ate devel�p�ent, but t� a regi�nally di��erent resp�nse �� the �ass balance at individual glaciers t� c��parable fluctuati�ns ��
�ete�r�l�gical para�eters caused by their di��erent glaci�l�gical characteristics.
In general, the net �ass balance at �ariti�e (western) glaciers in s�uthern ��rway is ��re in- fluenced by weather c�nditi�ns during the accu�u- lati�n seas�n than during the ablati�n seas�n, i.e.
�ainly by the a��unt �� winter precipitati�n/sn�w accu�ulati�n (Fig. 5). The c�rresp�nding net �ass balance at glaciers l�cated �arther east in a ��re c�ntinental cli�ate (St�rbreen, Hellstugubreen, Gråsubreen – all l�cated in J�tunhei�en) is exp�sed t� str�nger i�pacts �� variati�ns in e��ectiveness ��
su��er ablati�n and su��er balance (neSje et al.
2008a). On the basis �� this spatial di��erentiati�n between �ariti�e and c�ntinental glaciers in s�uth- ern ��rway, applicati�n �� large regi�nal glacier sa�ples like ‘Scandinavia’, ‘Eur�pe’, �r ‘Atlantic’
(e.g. IPCC 2007) �ay there��re be c�nsidered as ��
li�ited validity ��r specific appr�aches.
This study �� �ariti�e glaciers in s�uthern
��rway ��cuses in particular �n J�stedalsbreen and its �utlets. The �ass�balance series �� �igardsbreen (Fig. 6) is als� representative ��r the wh�le ice cap (c�. Winkler et al. 2009). Length�change
�easure�ents were per��r�ed at several �utlets (andreaSSen et al. 2005; c�. Fig. 7). The best c�rre- lati�n with the glacier variati�ns �� J�stedalsbreen has previ�usly been achieved with �ete�r�l�gi- cal data �r�� Bergen (lieStøl 1967; neSje 2005;
nOrdli et al. 2005). �i��erences between length changes �� individual �utlets �� J�stedalsbreen are related t� di��erences in �r�ntal ti�e lag depend- ent �n glacier t�ngue ge��etry, glacier size etc.
(neSje 1989; Winkler 1996; OerlemanS 2007).
Steep and sh�rt �utlets react �ast and are ��re sensitive t� changes in the �ass balance than �th-
ers. �ue t� its c�ntinu�us annual length variati�n rec�rd, Briksdalsbreen is ch�sen as an exa�ple here, whereas the �ther sh�rt �utlets sh�wed very si�ilar patterns �� length changes during the 20th century (Fig. 7b; c�. Winkler 1996). These highly sensitive reacting �utlets �� J�stedalsbreen under- went tw� c�ntrasting peri�ds �� str�ng advance
��ll�wed by rapid �r�ntal retreat during the past 20 years (Winkler et al. 2009). Interpretati�n �� this
‘extre�e’ behavi�ur deserves special attenti�n.
3 Length variations and mass balance dur- ing the past 20 years
An appr�xi�ately 30�year peri�d �� stati�nary
�r slightly advancing �r�ntal p�siti�ns at the sh�rt
�utlets �� J�stedalsbreen ended in the late 1980s.
Subsequently, these glaciers experienced a signifi- cant advance during the 1990s (Fig. 8). The cause
�� this advance was a c�nsiderable ice �ass gain,
�ainly achieved during a peri�d �� seven c�nsecu- tive p�sitive budget years �r�� 1989 t� 1996 (Figs.
3, 6a). The cu�ulative net �ass balance during that peri�d was +10.4 � w.e. (water equivalent) at
�igardsbreen (Winkler et al. 2009). Mete�r�l�gical data �r�� Bergen reveal a c�ncurrent trend �� in- creased winter precipitati�n (Tab. 2) c�rresp�nding t� the �ass increase and ab�ve�average winter bal- ances (andreaSSen et al. 2005; CHinn et al. 2005;
neSje et al. 2008a). A high c�rrelati�n between win- ter balances and b�th AO (Arctic Oscillati�n)� and
�AO (��rth Atlantic Oscillati�n)�indices indicates high cycl�nic activity and str�ng z�nal circulati�n with pred��inantly westerly airfl�w whilst the �ass gain took place (POHjOla and rOgerS 1997; neSje
et al. 2000; raSmuSSen and andreaSSen 2005).
There was neither a si�ilar increase in ice �ass, n�r an advance rec�rded at the c�ntinental glaciers in s�uthern ��rway (Winkler 2002; andreaSSen et al. 2005; c�. Figs. 2, 3).
At the end �� 1990s, the sh�rt �utlets ��
J�stedalsbreen entered a �ew years’ transiti�nal phase with stati�nary glacier �r�nts (Fig. 7a). With
�in�r l�cal deviati�ns, they started t� retreat ar�und the year 2000, rapidly accelerating t�wards the ��st recent years. The retreat at s��e glacier t�ngues led t� partial disintegrati�n (Fig. 9). It added up t� dis- tances exceeding the preceding advance at �any ��
the sh�rt �utlets. There��re, they presently �ccupy p�siti�ns well inside the 1980/90�ice li�it. The l�nger �utlets did j�in this ��st recent retreat with a delay �� several years. Until n�w, they have n�t
sh�wn a c��parable, �assive retreat (Winkler et al. 2009). The �act that �ariti�e glaciers in s�uth- ern ��rway n�w ��ll�w the ‘Gl�bal Trend’ (IPCC 2007; ZemP et al. 2008), a�ter receiving c�nsiderable attenti�n as excepti�ns just a �ew years ag�, needs t� be explained ��re care�ully.
4 Cause of the most recent length changes – a hypothesis
Acc�rding t� the net �ass�balance data �r��
�igardsbreen, there was a cu�ulative net �ass l�ss
�� c. 0.5 � w.e. �r�� 2000 t� 2008 (kjøllmOen Fig. 1: Mass-balance characteristics of six selected glaciers in southern Norway. (a) Means of seasonal and net mass-bal- ance values (for the period 1963–2008), (b) range of minimums, means, and maximums of seasonal and net mass-balance values (1963–2008), (c) maximum, mean, and minimum mass turnover (1963–2008), and (d) location map of investigated glaciers (raw data: NVE [ Norges Vassdrags- og Energidirektoratet ]; modified after Winkler 2009)
-2 000
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 -2 500
-1 500 -1 000 -500 m 0
Briksdalsbreen
Nigardsbreen Hellstugubreen
Storbreen
Fig. 2: Comparison of length changes (cumulative data) at Nigardsbreen, Briksdalsbreen (both Jostedalsbreen, maritime southern Norway), Storbreen, and Hellstugubreen (both Jotunheimen; raw data: NVE)
2009; Winkler et al. 2009). The length variati�ns during the ��st recent retreat at all the sh�rt �utlets see� n�t t� be pr�p�rti�nal t� this net �ass�balance rec�rd, especially i� c��pared t� the �agnitude ��
the �ass gain between 1988 and 2000 and its relat- ed advance (c�. Figs. 2, 3, 7; Tab. 6). Further��re, stati�nary glacier �r�nts were ��nit�red as early as 1997, and the retreat was already �ani�ested in 2000 at ��st �� the sh�rt �utlets (Fig. 7a). Acc�rding t�
previ�usly e�pirically calculated ter�inus reacti�n ti�es at th�se glaciers, varying between 3 and 4 years (neSje 1989, 2005; Winkler 1996), the retreat sh�uld have started later, i.e. a�ter c. 2004. Acc�rdingly, the
the�retically ��delled resp�nse ti�e �� 5 years ��r Briksdalsbreen (OerlemanS 2007) c�rresp�nded
�airly well with the real length change data �nly pri�r t� c. 1996. The cl�se link between the net �ass�bal- ance data rec�rd and length variati�ns evident since the early 1960s when �ass balance �easure�ents at
�igardsbreen started cann�t be detected a�ter 2000.
This phen��en�n is new at J�stedalsbreen, has n�t been rep�rted earlier, and is as yet n�t �ully under- st��d (see discussi�n).
Winter precipitati�n re�ained slightly ab�ve av- erage during the past 10 years (Winkler et al. 2009;
Tab. 2). A decrease in winter sn�w accu�ulati�n can,
Fig. 3: Net mass balances (cumulative data) of the glaciers from figure 1 (raw data: NVE)
1965 1970 1975 1980 1985 1990 1995 2000 2005
-20 -15 -10 - 5 0 5 10 15 20 m w.e.
Nigardsbreen
Ålfotbreen
Hardangerjøkulen
Storbreen
Hellstugubreen Gråsubreen
Table 1: Correlation (linear regression after Pearson) of (a) summer and (b) winter balances (annual data) of the six gla- ciers from figure 4. Despite the inherent variability in these annual data records, correlation is predominantly very posi- tive (raw data: NVE).
( r ) bs – bs Nigardsbreen Hardangerjøkulen Storbreen Hellstugubreen Gråsubreen
Ål��tbreen 0.70 0.70 0.75 0.71 0.64
�igardsbreen 0.90 0.86 0.85 0.73
Hardangerjøkulen 0.93 0.92 0.80
Storbreen 0.94 0.86
Hellstugubreen 0.88
(a)
( r ) bw - bw Nigardsbreen Hardangerjøkulen Storbreen Hellstugubreen Gråsubreen
Ål��tbreen 0.86 0.86 0.84 0.76 0.57
�igardsbreen 0.88 0.90 0.83 0.70
Hardangerjøkulen 0.82 0.75 0.52
Storbreen 0.91 0.77
Hellstugubreen 0.85
(b)
there��re, be ruled �ut as a cause �� the retreat. The
��st pr��inent �ete�r�l�gical signal is a rise �� air te�perature during the sec�nd hal� �� the ablati�n peri�d since the late 1990s (c�. Winkler et al. 2009;
Fig. 10). In �act, ��nthly �eans were up t� 2.0 °C ab�ve the n�r�al ��r 1961–1990 (Tab. 3). Already in the transiti�n peri�d, air te�peratures were c�nsid- erably higher in August and Septe�ber (c. 1.5 °C).
This c�rresp�nds t� stati�nary glacier t�ngues at
s��e �� the sh�rt �utlets. But whereas the rise in (late) su��er air te�peratures clearly c�rrelates and explains the length changes, th�se length changes still c�nflict with the net �ass�balance data.
In the�ry, each glacier t�ngue sh�uld resp�nd with a specific �r�ntal ti�e lag (here re�erred t� as ‘ter-
�inus reacti�n ti�e’) t� any deviati�n �r�� a steady�
state �ass flux (Fig. 11). In the present situati�n, this dyna�ic resp�nse see�s �bvi�usly ‘disturbed’ as
Fig. 4: Annual deviation of (a) summer and (b) winter balances from the individual 1963–2008 means (set as 100 %) of the glaciers from figure 1 (cf. Tab. 1; raw data: NVE)
Fig. 5: (a) Correlation (linear regression after Pearson) of seasonal balances with net balance (annual data) and (b) re- lationship between related standard deviations of the means for the period 1963–2008 for the glacier from figure 1. The corresponding data for other selected periods are given for comparison. For detailed discussion of these relations/param- eters see e.g. nesje et al. 2000; rasmussen and ConWay 2001; Winkler 2002; andreassen et al. 2005 (raw data: NVE)
0.9 0.8 0.7 0.6 0.5 0.4
σbw/σbn σbs/σbn
1963-2008 0.84 0.85 0.86 0.74 0.64 0.57 0.67 0.83 0.71 0.85 0.90 0.92 1963-2000 0.86 0.85 0.85 0.75 0.60 0.54 0.55 0.82 0.69 0.79 0.88 0.88 1963-1996 0.88 0.86 0.89 0.80 0.67 0.58 0.59 0.82 0.72 0.79 0.86 0.85 1989-1996 0.93 0.97 0.95 0.94 0.91 0.75 0.31 0.78 0.61 0.80 0.76 0.76 2001-2008 0.83 0.85 0.89 0.72 0.69 0.72 0.80 0.84 0.72 0.89 0.92 0.98
0.75 0.61 0.72 0.55 0.44 0.41 0.56 0.58 0.53 0.71 0.79 0.84 0.84 0.63 0.74 0.62 0.48 0.47 0.51 0.57 0.54 0.70 0.81 0.85 0.81 0.62 0.73 0.62 0.52 0.53 0.48 0.57 0.49 0.65 0.76 0.82 0.95 0.78 0.84 0.71 0.72 0.65 0.38 0.32 0.32 0.42 0.47 0.67 0.63 0.59 0.74 0.47 0.42 0.24 0.59 0.59 0.47 0.74 0.77 0.84 1.0
0.9 0.8 0.7 0.6
0.5 Ålf. Nig. Ha. Sto. He. Gr. Ålf. Nig. Ha. Sto. He. Gr.
bw - bn bs - bn
a) r b)
Nigardsbreen
Ålfotbreen Hardangerjøkulen Hellstugubreen Storbreen Gråsubreen
1965 1975 1985 1995 2005
0 50 100 150 200 %
1965 1975 1985 1995 2005
50 %
100 150 200
0
bw - øbw bs - øbs
length variati�ns since c. 1997 �nly partially reflect the net �ass�balance changes (Fig. 12). The c��bi- nati�n �� steep, �ast�reacting �utlets with a �ariti�e glaci�l�gical regi�e with high �ass turn�ver and steep �ass�balance gradients �ake these glaciers ex- tre�ely sensitive and c��plex in their resp�nse t�
cli�ate changes (Winkler 2009). One result is c�n- siderable seas�nal length variati�ns. Except in years with rapidly advancing glacier �r�nts, su��er �elt- back always causes reas�nable seas�nal retreat �� the l�w�lying glacier �r�nts (Winkler 2008, 2009). I�
su��er �eltback is enhanced due t� extre�ely high su��er air te�peratures and extra�rdinary ablati�n rates, this will a��ect the glacier t�ngue i��ediately with�ut any �r�ntal ti�e lag (Fig. 11). Further��re, this reacti�n �ight be largely independent �� the net
�ass�balance situati�n �� the glacier as a wh�le. F�r exa�ple, increasing �r decreasing winter sn�w accu-
�ulati�n will pri�arily be e��ective in the accu�ula- ti�n area �� the glacier and, there��re, deter�ined by the triggered changes �� the �ass flux sh�w its
influence �n the glacier ter�inus (with the specific ti�e lag). But even in the status �� a ‘n�r�al’ �r even slightly enhanced �ass flux due t� preceding bal- anced �r slightly p�sitive budget years, the reacti�n
�� the ter�inus t� the �ass flux can be disturbed by extra�rdinary su��er �eltback. Thus, a high an- nual retreat �ight be �easured despite a lack �� a related expected signal in the net balance data.
The steep �utlets �� J�stedalsbreen are rare ex- a�ples �� significant changes in ter�inus reacti�n ti�es. Alth�ugh separate ter�inus reacti�n ti�es related t� precipitati�n and air te�perature have pre- vi�usly been suggested (Salinger et al. 1983), such suggesti�ns have n�t been pursued in the atte�pts t�
calculate resp�nse ti�es �� glaciers �n a the�retical level in c��binati�n with ��delling (jóHanneSSOn
et al. 1989; leySinger Vieli and gudmundSSOn
2004; HOOke 2005; marSHall 2006; OerlemanS 2007). In the case presented here, individual extents related t� di��erent c�nditi�ns during the ��st re- cent advance and retreat peri�ds have t� be applied.
b) Nigardsbreen
-12 -10 -8 -6 -4 -2 0 2
m w.e. bn 4 1 000
500 1 500 m a.s.l. Altitude
2006 1989
2 000
Nigardsbreen a)
-1.0 0
m w.e. Net balance
Winter balance
Summer balance
1.0 2.0 3.0 -3.0 -2.0 -1.0 0 m w.e.
m w.e.
0 1.0 2.0 3.0 4.0
1965 1970 1975 1980 1985 1990 1995 2000 2005
Fig. 6: (a) Mass-balance data for Nigardsbreen and (b) net mass-balance gradients for Nigardsbreen during the period 1987–2008, both showing considerable annual variations. 1989 was the year with the most positive, 2006 with the most negative net balance (data: NVE; modified after Winkler 2009)
As l�ng as winter precipitati�n is the pred��inant cli�at�l�gical para�eter, the glacier t�ngue will re- sp�nd dyna�ically t� changes �� the �ass flux and the previ�usly c�nfir�ed ter�inus reacti�n ti�e (e.g. 3 t� 4 years at Briksdalsbreen) will fit per�ectly.
I� excessive ablati�n caused by extre�ely high su�-
�er air te�peratures adds t� the �ass l�ss at the glacier t�ngue, ter�inus reacti�n ti�es will dr�p t�
zer� and the glacier �r�nt will react instantane�usly t� the weather c�nditi�ns during the ablati�n sea- s�n. This str�ng influence �� additi�nal enhanced
�eltback disturbing the pure dyna�ic resp�nse t�
changes �� the �ass flux is a new finding. As t� �ur kn�wledge, this is n�t yet i�ple�ented in the�reti- cal ��dels and related studies �n the resp�nse ti�es
�� glaciers.
10 km N Jostedalsbreen
Jostedalen Lodalen
Oldedalen
Myklebustbreen
1
1 Austerdalsbreen 2 Bergsetbreen 3 Bødalsbreen 4 Bøyabreen 5 Brenndalsbreen 6 Briksdalsbreen 7 Fåbergstølsbreen
8 Kjenndalsbreen 9 Melkevollbreen 10 Nigardsbreen 11 Stegholtbreen 12 Supphellebreen 13 Tuftebreen 14 Tunsbergdalsbreen
2 3
4 56 9
7 8
10 10 13
11
14
12 0
a)
1996 1998 2000 2002 2004 2006 2008
-600 -500 -400 -300 -200 -100 0 m
Briksdalsbreen Briksdalsbreen
Kjenndalsbreen Kjenndalsbreen Brenndalsbreen Brenndalsbreen
Bergsetbreen Bergsetbreen Bødalsbreen
Bødalsbreen c)
b)
a)
Jostedalsbreen Short outlets
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000
length change (100 m-intervals)
Briksdalsbreen Briksdalsbreen
Brenndalsbreen Bergsetbreen
Bødalsbreen Kjenndalsbreen Kjenndalsbreen Bøyabreen Supphellebreen Melkevollbreen
Fig. 7: (a) Most recent length variations of selected short outlets of Jostedalsbreen. Due to morphological changes of the glacier tongue the series at Bergsetbreen stopped in 2006 (cf. Winkler et al. 2009). (b) Length variations at short outlets of Jostedalsbreen. Except for Briksdalsbreen, all measurements stopped around 1950. Except for Melkevollbreen, the measurements resumed in the 1990s (starting point not in scale to retreat in intermittent period). (c) Location map of Jostedalsbreen and its outlets (raw data: NVE; S. Winkler)
Phase: I “Advance” (1989–
1995) I/II “Transiti�n”
(1996–2000) II “Retreat” (2001–
2008) Normal (1961/1990) Σ precipitati�n
October 166 �� 318 �� 236 �� 271 ��
��ve�ber 217 �� 197 �� 328 �� 259 ��
�ece�ber 341 �� 232 �� 288 �� 235 ��
January 334 �� 219 �� 289 �� 190 ��
February 234 �� 329 �� 182 �� 157 ��
March 301 �� 210 �� 155 �� 170 ��
April 165 �� 124 �� 145 �� 114 ��
Table 2: Deviation of monthly winter precipitation from the 1961–90 normals for Bergen during the past 20 years, calcu- lated as means for three periods representing advance, transition and retreat at the short outlets (raw data: DNMI [ Det Norske Meteorologiske Institutt ]/Met.no).
a)
b)
c)
07.08.1989 21.05.1998
25.06.1996 11.08.1990
22.05.1998 22.06.1996
05.06.1994 08.08.1990
Fig. 8: Visual comparison of morphological changes at the glacier tongues of (a) Briksdalsbreen, (b) Supphellebreen, and (c) Bøyabreen during the 1990s advance (all photos © by s. Winkler; complete photo series of 12 outlets of Jostedalsbreen are available at: http://www.geographie.uni-wuerzburg.de/arbeitsbereiche/physische_geographie/weitere_forschung- sarbeiten/norglamo)
a)
b)
c)
05.09.2001 27.08.2008
24.06.2000 01.09.2004 29.08.2008
08.09.2001 23.06.2005 29.08.2008
Fig. 9: Visual comparison of morphological changes at the glacier tongues of (a) Briksdalsbreen, (b) Kjenndalsbreen, and (c) Melkevollbreen (all photos © by S. Winkler)
We sh�uld �enti�n �ne �eth�d�l�gical pr�b- le� here. �ue t� accessibility and �ther technical reas�ns the altitudinal values �� the net �ass bal- ance ��r the l�wer��st 200–300 � �� the glacier sur�ace �� �igardsbreen usually are extrap�lated (kjøllmOen 2008). In years with high su��er ablati�n rati�s, the c�nventi�nal data �ight there-
��re t� a certain degree underesti�ate the net
�ass l�ss at the l�wer��st t�ngue (Winkler et al.
2009). �ue t� the relative s�all percentage �� gla- cier area and v�lu�e l�cated at this l�w altitude, the degree �� uncertainty see�s t� be acceptable in the light �� the t�tal glacier v�lu�e. H�wever, especially ��r the narr�w and steep t�ngue �� the sh�rt �utlets, the c�nsequences ��r t� explain the length changes cann�t be �verseen. On the �ther hand, th�se p�tential �eth�d�l�gical pr�ble�s d� n�t �ully explain the present situati�n, as pri�r t� 2000 they see� n�t t� have had any c�nsider- able i�pact �n the relati�nship between net �ass�
balance and length�change data. Further��re, l�- cal influences can ��st likely be ruled �ut due t�
the parallel trend �ccurring at all sh�rt �utlets ��
J�stedalsbreen (Figs. 7a, 9). Even i�, ��r exa�ple,
calving �ver a s�all pr�glacial lake influenced the
�ast retreat �� Briksdalsbreen and is assu�ed as re- sp�nsible ��r the n�n�repr�ducibility in an up�t��
date si�ulati�n (laumann and neSje 2009), this explanati�n cann�t be applied at �ther �utlets with a c��parable str�ng retreat.
5 Changes of the glaciological regime?
There are clear indicati�ns that the glaci�- l�gical regi�e was partially ��dified a�ter 2000.
Substantial changes �ccurred with the c�rrelati�n
�� the net �ass�balance para�eters t� length vari- ati�ns (Tab. 4). The c�rrelati�n �� net �ass balance t� length changes dr�pped significant during the
��st recent retreat since 2000 indicating that the previ�us cl�se c�upling �� net �ass�balance data rec�rd and length variati�ns cann�t l�nger be ap- plied unaltered t� explain the �r�ntal behavi�ur.
Further��re, it can be detected that the c�nsidera- ti�n �� ter�inus reacti�n ti�es n�w leads t� less significant results. This �eans that length changes at the ter�inus are n�w less influenced by the �ass trans�er in �av�ur �� an i��ediate resp�nse t� (ex- cessive) su��er ablati�n. It als� explains why (n�t delayed) su��er balance data experiences a high c�rrelati�n with length changes (r = 0.97) during the last �ew years (Winkler et al. 2009) despite ��
its �eth�d�l�gical uncertainties and partial de- pendency �n the preceding winter balance. The negative c�rrelati�n �� net �ass balance with �r�n- tal variati�ns during the transiti�n peri�d (�ean- ing si�ultane�us �ass increase and �r�ntal retreat) see�s striking.
C��parable changes between l�ng�ter�
�eans and the ��st recent retreat phase take place between selected �ete�r�l�gical para�eters and length variati�ns (Tab. 5). The high c�rrelati�n
�� winter te�peratures with length changes, eas- ily explained by the i�pact �� winter precipitati�n
1970 1975 1980 1985 1990 1995 2000 2005
70 60 50 40 30 20 10 0 -10 -20
°C
Σ (δ T7 - δ T9)
Bergen
Σ (δ T5 - δ T9)
Fig. 10: Cumulative deviations of summer air temperatures for Bergen (annual sum of monthly deviations May to Sep- tember and July to September) from the related 1961–90 nor- mals (raw data: DNMI/Met.no)
Phase: I “Advance”
(1989–1995) I/II “Transiti�n”
(1996–2000) II “Retreat”
(2001–2008) Normal
(1961/1990)
∅ air te�perature
May 10.4 °C 10.0 °C 10.7 °C 10.5 °C
June 12.9 °C 13.0 °C 13.7 °C 13.3 °C
July 14.7 °C 14.5 °C 16.3 °C 14.3 °C
August 14.1 °C 15.6 °C 16.1 °C 14.1 °C
Septe�ber 11.6 °C 12.7 °C 12.7 °C 11.2 °C
Table 3: Deviation of monthly summer air temperatures from the 1961–90 normals for Bergen, calculations corresponding to table 2 (raw data: DNMI/Met.no).
�n the �ass balance (�ild winter = high sn�w ac- cu�ulati�n), has dr�pped significantly during the
��st recent years. By c�ntrast, the results ��r su�-
�er air te�perature para�eters (p�sitive depar- tures c�rresp�nd t� negative length changes) have changed as expected. I� c��pared t� the values ��r the preceding peri�d, it bec��es �bvi�us that su�-
�er air te�perature n�wadays has a higher i�pact
�n length changes than be��re.
This analysis reveals substantial changes in the relati�nships and interrelati�n within the glaci�- l�gical regi�e during the past 20 years. The par- tially surprising results give a clear warning t� the c����n pr�cedure �� averaging l�ng�ter� data series �� glaci�l�gical and �ete�r�l�gical para�- eters as input ��r existing ��dels. These inputs
�ust n�t aut��atically be c�nsidered as c�nstant.
Te�p�rally and spatially di��erentiated, high�res-
�luti�n data and �ultiple�phase regressi�n see�
t� be appr�priate strategies in the c�ntext �� sig- nificant weather regi�e changes. H�wever, this re-
�ers �erely t� length variati�n data, as n� distinct regi�e change was detected in the �ass�balance data (Fig. 5). Only during the sh�rt peri�d �r��
2001 t� 2006, a �ew c�e�ficients sh�wed s��e
��derate departures, e.g. at Ål��tbreen (bs–bn: r
= 0.66 ��r 2001–2006; r = 0.80 ��r 2001–2008) �r
�igardsbreen (bw–bn: r = 0.71 ��r 2001–2006; r = 0.85 ��r 2001–2008).
6 Annual air temperatures indicative for length changes?
In s��e studies, glacier length variati�n data is used t� calculate the air te�perature rise during the 20th century and deduce anthr�p�genic influences (e.g. OerlemanS 1994, 2005). Detailed glaciological and �ete�r�l�gical data �r�� s�uthern ��rway al�ng with general c�nsiderati�ns (c�. Winkler 2002) raise, h�wever, �aj�r c�ncerns with that practice. Firstly, annual air te�perature data are n�r�ally related t�
calendar years. By c�ntrast, glacier �ass�balance data are calculated ��r glacier budget years (1st October
- ‘normal’ - - ‘disturbed’ -
mass gain
Dynamic response
terminusreaction time
excessive summer backmelting
mass gain
mass loss mass loss
length change
positivenegative advanceretreat
net balance
Fig. 11: Schematic visualisation of the response of a glacier tongue to mass flux fluctuations (cf. text)
Fig. 12: Comparison of length variations of Briksdalsbreen with the cumulative net mass-balance series of Nigards- breen (raw data: NVE)
1980 1985 1990 1995 2000 2005
-300 -200 -100 0 100 200 300 m
Briksdalsbreen
(length change) Nigardsbreen (net balance)
m w.e.
-15 -10 -5 0 5 10 15
to 30th Septe�ber). Annual length changes are reg- ularly �easured in late su��er and r�ughly c�rre- sp�nd t� budget years. When c��paring ‘annual’ air te�peratures ��r Bergen calculated ��r b�th calendar years and budget years, di��erences �� up t� ±1°C e�erged (Fig. 13a). The average deviati�n ��r the peri�d 1900–2000 was ±0.28 °C and increased t�- wards the ��st recent years (Fig. 13b). Only a rather weak trend was derived lacking any clear pattern �r c�nspicu�us explanati�n. As a c�nsequence, i� �ete-
�r�l�gical para�eters �n an annual basis are linked t� �ass�balance �r length variati�n data, they �ust be calculated ��r budget years and n�t ��r calendar years. This rather l�gical rule has n�t yet received enough attention.
There see�s t� be n� �bvi�us relati�n �� net
�ass�balance �r length change data �r�� �ariti�e s�uthern ��rway during the past decades t� the av- eraged �ete�r�l�gical para�eter ‘annual air te�pera- ture’, even i� the latter �ne is calculated ��r budget years (see Tab. 6). By c�ntrast, alth�ugh annual air te�peratures are ab�ve average during the ��st re- cent years with its str�ng �r�ntal retreat and (�in�r) net �ass l�ss, they were als� ab�ve the l�ng�ter�
�eans during the 1990s with its c�nsiderable net
�ass increase and advance. The latter phen��en�n was due t� high winter air te�peratures (Fig. 13c).
Relatively high winter air te�peratures indicate �ild and ��ist winter seas�ns with c�nsiderable sn�w ac- cu�ulati�n and, there��re, �av�urable c�nditi�ns ��r the glaciers. By c�ntrast, ab�ve average su��er air te�peratures cause high ablati�n and high �ass l�ss in su��er. P�sitive departures �� the air te�perature
�r�� the l�ng�ter� �eans during the ablati�n seas�n have, there��re, a very negative i�pact �n the glaciers.
�egative departures �� air te�peratures e.g. during c�ld and dry winter seas�ns are less �av�urable ��r the net �ass balance, and l�w su��er air te�peratures will reduce ablati�n in �av�ur �� a p�sitive net bal- ance. F�r s�uthern ��rway, these c�nsequences are c�nfir�ed by e�pirical data e.g. ��r the ‘Little Ice Age’
(neSje and daHl 2003; neSje et al. 2008b).
Alth�ugh air te�peratures have an i�p�rtant i�- pact �n the net �ass balance and length variati�ns at the �ariti�e glaciers �� s�uthern ��rway, they need, there��re, t� be analysed in the light �� seas�nality.
Mid�latitudinal �ariti�e glaciers experience a �arked seas�nal di��erentiati�n in their �ass budget. The di�- ( r ) length change
Briksdalsbreen: Phase I
“Advance”
(1989–1995)
Phase I/II
“Transiti�n”
(1996–2000)
Phase II
“Retreat”
(2001–2008)
Peri�d
1963–2007 Peri�d
1963–2000 Peri�d 1963–1996
Net balance 0.96 -0.94 0.50 0.77 0.93 0.89
Net balance (tr: 3 a) 0.94 -0.93 n/a 0.70* 0.79 0.92
Net balance (tr: 4 a) 0.85 -0.96 n/a 0.64* 0.88 0.94
Net balance (tr: 5 a) 0.47 -0.92 n/a 0.57* 0.64 0.92
(* � peri�d ends 2002 t� 2004 acc�rding t� particular ter�inus resp�nse ti�e)
Table 4: Correlation (linear regression after Pearson) of different net mass-balance parameters for Nigardsbreen with length variations at Briksdalsbreen (cumulative data series). The three multi-year phases correspond to those in table 2.
For net balance, length changes with different terminus reaction times of 3, 4, and 5 years [tr: 3a, 4a, 5a] are included (raw data: NVE; in parts modified after Winkler et al. 2009).
( r ) length change
Briksdalsbreen: Phase I
“Advance”
(1989–1995)
Phase I/II
“Transiti�n”
(1996–2000)
Phase II
“Retreat”
(2001–2008)
Peri�d
1963–2007 Peri�d
1963–2000 Peri�d 1963–1996
Σ δ(T Oct–Apr) 0.76 -0.91 -0.91 0.60 0.90 0.86
Σ δ(T Jun–Sep) 0.63 -0.70 -0.96 0.27 0.81 0.73
Σ δ(T May–Sep) 0.37 -0.65 -0.93 0.14 0.55 0.09
Σ δ(T May–Sep) (tr: 3a) 0.66 -0.91 n/a -0.06* 0.37 0.05 Σ δ(T May–Sep) (tr: 4a) 0.78 -0.72 n/a -0.10* 0.28 0.07 Σ δ(T May–Sep) (tr: 5a) 0.61 -0.83 n/a -0.11* 0.17 0.08
(* � peri�d ends 2002 t� 2004 acc�rding t� particular ter�inus resp�nse ti�e)
Table 5: Correlation (linear regression after Pearson) of different air temperature parameters (sum of deviation of monthly air temperature means from 1961–90 normals) for Bergen with length variations at Briksdalsbreen (cumulative data se- ries). The three multi-year phases correspond to those in table 2. For the sum of May–September, length changes with different terminus response times of 3, 4, 5 years [tr: 3a, 4a, 5a] are included (raw data: DNMI/Met.no, NVE; in parts modified from Winkler et al. 2009).
�erentiati�n between winter and su��er balance is regarded as use�ul ��r the analysis �� the influence ��
single cli�at�l�gical �act�rs up�n the �ass�balance (dyurgerOV and meier 1999; Steiner et al. 2008).
This indicates why para�eters that d� n�t include such seas�nal di��erentiati�n, as ��r exa�ple, annual air te�perature data, are p��r indicat�rs �� glacier
�ass balances �r related length variati�ns in �ariti�e
�id�latitude envir�n�ents. A success�ul applicati�n
�� this para�eter w�uld require that (a) �nly the cli-
�atic c�nditi�ns during the ablati�n seas�n deter�ine the net �ass balance and (b) �nly su��er air te�per- atures sh�w detectable departures �r�� the l�ng�ter�
�eans. As (a) already has been dispr�ved ��r s�uthern
��rway (Fig. 5), (b) is �airly unrealistic and re�utable
�n the basis �� existing data as well (Fig. 13).
As n� causal link between the para�eter ‘annual air te�perature’ and glacier variati�ns c�uld be de- tected during the past decades in s�uthern ��rway, these �ariti�e glaciers sh�uld be excluded �r�� the ab�ve��enti�ned studies, at least �n decadal� and secular�ti�e scales. Air te�perature data need t� be seas�nally di��erentiated and c��bined with precipita- ti�n and eventually �ther energy balance variables be-
��re it �ight success�ully be related t� length changes.
This rec���endati�n �ight c�nsiderably c��plicate
the existing atte�pts t� extract a single cli�atic signal
�r�� the available length change data available, but see�s necessary in �rder t� pr�ve that, ��r exa�ple,
‘parallel’ trends �� rising annual air te�peratures and glacier retreat during the 20th century (��ll�wing the
‘Little Ice Age’) are �� ��re than statistical nature. At least ��r �ariti�e �id�latitude ��untain glaciers, the causal links between annual air te�perature data sets (especially i� they are averaged �n a gl�bal scale) and length variati�n are still awaiting ulti�ate pr��� ��
linkages.
7 Conclusions and outlook
In su��ary, s��e �acts ab�ut the ��st recent glacier variati�ns in s�uthern ��rway need t� be ad- dressed, regarding the �requent use �� data ��r ��d- elling and related studies. Since 2000, length varia- ti�ns at the sh�rt �utlets �� J�stedalsbreen see� t�
be dec�upled �r�� the net �ass�balance data series.
The dyna�ic resp�nse �� the glacier �r�nt t� net bal- ance and �ass flux variati�ns has been disturbed.
Previ�usly applicable ter�inus reacti�n ti�es �� 3 t�
4 years have been replaced by an i��ediate resp�nse t� higher su��er air te�peratures. Even i� their de-
ø (T5 - T9) ø (T10 - T 9)
ø (T10 - T 4)
ø (T1 - T12; cal.yr.) ø (T10 - T9; bud.yr.)
b)
a) (Tcal.yr. - Tbud.yr.) c)
Bergen
°C
°C ° C
9.5 9.0 8.5 8.0 7.5 7.0 6.5 -1.25 -1.0 -0.75 -0.25 0.25 0.75 1.0
0.50
0
-0.5
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000
1970 1975 1980 1985 1990 1995 2000 2005 21970 1975 1980 1985 1990 1995 2000 2005
3 4 5 6 7 8 9 10 11 12 13 14 15
Fig. 13: (a) Difference between ‘annual air temperatures’ related to calendar and budget years, respectively, for Bergen. (b) Comparison of ‘annual air temperatures’ related to calendar and budget years, respectively, for Bergen. The solid lines are five-year Gaussian low-pass filters. (c) Comparison between winter, summer, and (budget year) annual air temperatures for Bergen. The solid lines are five-year Gaussian low-pass filters (raw data: DNMI/Met.no)
gree �� d��inance underwent a slight decrease ��l- l�wing the 1990s advance, winter balance and winter precipitati�n c�nditi�ns are still i�p�rtant driving �ac- t�r ��r the �ass�balance. A way �� integrating p�s- sible regi�e shi�ts int� glacier ��dels needs t� be deter�ined. This is essential as relatively sh�rt ti�e peri�ds experience substantial �r�ntal advance and retreat. Because e�pirical explanati�ns and ��dels based �n l�ng�ter� data that previ�usly delivered sat- is�act�ry s�luti�ns ��r the interpretati�n �� the glacier variati�ns d� n�t w�rk in the present situati�n (since 2000), they sh�uld n�t be applied untested ��r ��del- ling �uture glacier variati�ns. On the �ther hand, ��d- els based �n �easure�ents per��r�ed during the �ew years a�ter 2000 will n�t necessarily be representative
��r glacier behavi�ur be��re 2000. That a �aj�r retreat like the �ne since 2000 �ight �ccur with�ut a si�ulta- ne�us net �ass l�ss needs t� be taken int� acc�unt i�
glacier variati�ns ��r peri�ds pri�r t� the start �� �ass balance ��nit�ring are analysed.
As it is crucial t� understand the resp�nse �� the glacier t�ngues t� �uture cli�ate change, sh�rt�ter�
extre�e situati�ns – rather than l�ng�ter� c�nstant devel�p�ents – need t� be the ��cus �� new studies.
The c�ncept �� p�ssible regi�e shi�ts and �ulti�
phase patterns �� causal interacti�ns between �ete-
�r�l�gical para�eters and glacier length changes has t� be accepted as hyp�thesis in �uture w�rk.
Finally, the pure dyna�ic resp�nse �� the glacier
�r�nt t� �ass�balance changes can be disturbed at l�w�lying �ariti�e ��untain glaciers with high rates
�� su��er �eltback. Alterati�n �� ter�inus reac- ti�ns ti�es can be used as indicat�r. I� the ter�inus resp�nse is �ainly driven by changes �� the �ass flux, a specific ter�inus reacti�n ti�e (tr ≥ 1 yr.) will be in place. I� enhanced ablati�n at the ter�inus cr�ssed a specific thresh�ld, this ter�inus reacti�n ti�e will be replaced by an i��ediate resp�nse (tr
= 0) independent �� the �ass flux deter�ined by the net �ass�balance situati�n during the preceding
years. Whereas �ass balance and dyna�ic resp�nse
�� the glacier �r�nt �ight be ��delled with existing kn�wledge and available c�nventi�nal �ass�balance data, t� quanti�y these surplus ablati�n/additi�nal enhanced su��er �eltbacks �ight be a �eth�d-
�l�gical task ��r the �uture. This c�uld be under- taken by applying new pr�cedures and techniques (e.g. high�res�luti�n ge�detic �eth�ds like annually repeated terrestrial laser scanning). Si�ultane�usly, thresh�lds �� su��er air te�perature specific ��r this regi�n need t� be ��und in �rder t� resp�nd t� this relatively new devel�p�ent that is with�ut anal�gy in the hist�ric �r recent rec�rds. As a c�n- sequence, di��erent ter�inus reacti�n ti�es ��r ad- vance/retreat, precipitati�n/air te�perature, �r dy- na�ic resp�nse/disturbed dyna�ic resp�nse have t� be intr�duced. In the light �� cli�ate ��dels unani��usly ��recasting an increase in precipita- ti�n in �ariti�e s�uthern ��rway (beldring et al.
2007; Haugen et al. 2008; SOrteberg and anderSen
2008), all types �� glacier ��dels including precipi- tati�n as a c�nstant input (e.g. certain energy�bal- ance ��dels and their derivates) will pr�duce results that have pri�arily t� be regarded as less realistic.
H�wever, it will be an i�p�rtant questi�n whether winter precipitati�n retains its str�ng influence dur- ing the 21st century and length variati�ns re�ain dec�upled �r�� the c�nventi�nal net �ass�balance rec�rd, thus c��plicating any atte�pt t� ��recast th�se length changes.
Acknowledgements
We thank the c�lleagues �r�� the ��rwegian Water Res�urces and Energy �irect�rate (NVE)
��r l�ng and �ruit�ul c��perati�n. Martin Miles (University �� Bergen) and Martin Br��k (Massey University Pal�erst�n ��rth) kindly i�pr�ved the language. SW wants t� ackn�wledge financial sup- Period ∅ Annual air te�perature
(budget year) Bergen Σ Net balance
�igardsbreen Σ Length change Briksdalsbreen
2001–2007 8.56 °C +0.15 � w.e. �408 �
1996–2000 8.02 °C +2.49 � w.e. 0 �
1989–1995 8.10 °C +10.37 � w.e. +230 �
1981–1988 7.53 °C +1.71 � w.e. �9 �
1970–1980 7.70 °C +0.96 � w.e. +132 �
Table 6: Comparison between mean annual air temperatures for Bergen (calculated for budget years), net mass balance for Nigardsbreen (cumulative date), and length changes at Briksdalsbreen (cumulative data) for selected periods. This table clearly shows that there is no discernible pattern between the data series, raising serious doubts whether a causal link between length-change records and net mass balance/annual air temperatures really can be concluded for this re- gion (raw data: DNMI/Met.no, NVE).
p�rt �� the �eutsche F�rschungsge�einscha�t (�FG�Grant: WI 1701/3). Al Ras�ussen and tw�
an�ny��us reviewers gave valuable c���ents t� an earlier versi�n �� this paper.
References
andreaSSen, l. m.; elVeHøy, H.; kjøllmOen, b.; engeSet, r. V. and HaakenSen, n. (2005): Glacier �ass balance and length variati�ns in ��rway. In: Annals �� Glaci�l- ogy 42, 317–325. Doi:10.3189/172756405781812826 bamber, j. l. and Payne, a. j. (eds.) (2004): Mass balance ��
the cry�sphere. Observati�ns and ��delling c�nte�p�- rary and �uture changes. Ca�bridge.
beldring, S.; rOald, l. a.; engen-Skaugen, t. and før-
land, e. j. (2007): S�å endringer kann �å st�r betydning.
In: Kli�a 2, 32–34.
CHinn, t. j. H.; Winkler, S.; Salinger, m. j. and Haaken-
Sen, n. (2005): Recent glacier advances in ��rway and
�ew Zealand – a c��paris�n ��r their glaci�l�gical and
�ete�r�l�gical causes. In: Ge�grafiska Annaler 87 A, 141–157.
dyurgerOV, m. b. and meier, m. f. (1999): Analysis �� win- ter and su��er glacier �ass balance. In: Ge�grafiska Annaler 81 A, 541–554.
Haugen, j. e.; køltZOW, m. and iVerSen, t. (2008): Mer ekstre� nedbør �g vind i ��rge. In: Kli�a 2, 36–38 HOOke, r. l. (20052): Principles �� glacier �echanics. Ca�-
bridge.
Huber, u. m.; reaSOner, m. a. and bugmann, H. (eds.) (2005): Gl�bal change and ��untain regi�ns – a state ��
knowledge overview. Dordrecht.
IPCC (ed.) (2007): Cli�ate change 2007: the physical science basis. C�ntributi�n �� w�rking gr�up I t� the 4th Assess-
�ent rep�rt �� the Interg�vern�ental Panel �n Cli�ate Change. Ca�bridge.
jóHanneSSOn, t.; raymOnd, C. f. and WaddingtOn, e. d.
(1989): Ti�e scale ��r adjust�ent �� glaciers t� changes in �ass balance. In: J�urnal �� Glaci�l�gy 35, 355–369.
kjøllmOen, b. (ed.) (2008): Glaci�l�gical investigati�n in Norway 2007. �VE Rapp�rt 2008/1. Osl�.
– (ed.) (2009): Glaci�l�gical investigati�n in ��rway 2008.
�VE Rapp�rt 2009/2. Osl�.
laumann, t. and neSje, a. (2009): A si�ple �eth�d ��
si�ulating the �uture �r�ntal p�siti�n �� Briksdals- breen, western ��rway. In: The H�l�cene 19 (in press).
Doi:10.1177/0959683608100566
leySinger Vieli, g. j. m. C. and gudmundSSOn, g. H.
(2004): On esti�ating length fluctuati�ns �� gla- ciers caused by changes in cli�atic ��rcing. In: J�ur- nal �� Ge�physical Research 109 (F1), F01007.
��i:10.1029/2003JF000027.
lieStøl, O. (1967): St�rbreen glacier in J�tunhei�en.
��rsk P�larinstitutt Skri�ter 141. Osl�.
marSHall, S. j. (2006): M�delling glacier resp�nse t�
cli�ate change. In: knigHt, P. g. (ed.) (2006): Gla�(2006): Gla- cier science and envir�n�ental change. Ox��rd, 163–173.
neSje, a. (1989): Glacier��r�nt variati�ns at the �utlet glaciers �r�� J�stedalsbreen and cli�ate in the J�s- tedalsbre regi�n �� western ��rway in the peri�d 1901–1980. In: ��rsk Ge�grafisk Tidsskri�t 43, 3–17.
Doi:10.1080/00291958908552213
– (2005): Briksdalsbreen in western ��rway: A�
1900–2004 �r�ntal fluctuati�ns as a c��bined e�-
�ect �� variati�ns in winter precipitati�n and su�-
�er te�perature. In: The H�l�cene 15, 1245–1252.
Doi:10.1191/0959683605hl897rr
neSje, a. and daHl, S. O. (2003): The ‘Little Ice Age’
– �nly te�perature? In: The H�l�cene 13, 139–145.
��i:10.1191/0959683603hl603�a
neSje, a.; lie, ø. and daHl, S. O. (2000): Is the ��rth Atlantic Oscillati�n reflected in Scandinavian gla- cier �ass balance rec�rds? In: J�urnal �� Qua- ternary Science 15, 587–601. Doi:10.1002/1099- 1417(200009)15:6<587::AI��JQS533>3.0.CO;2�2 neSje, a.; bakke, j.; daHl, S. O.; lie, ø. and mattHeWS,
j. a. (2008a): ��rwegian ��untain glaciers in the past, present and �uture. In: Gl�bal and Planetary Change 60, 10–27. ��i:10.1016/j.gl�placha.2006.08.004 neSje, a.; daHl, S. O.; tHun, t. and nOrdli, ø. (2008b):
The “Little Ice Age” glacial expansi�n in western Scandinavia: su��er te�perature �r winter precipita- ti�n? In: Cli�ate �yna�ics 30, 789–801. Doi:10.1007/
s00382�007�0324�z
nOrdli, ø.; lie, ø.; neSje, a. and beneStad, r. (2005):
Glacier �ass balance in s�uthern ��rway ��delled by circulati�n indices and spring�su��er te�pera- ture A� 1781–2000. In: Ge�grafiska Annaler 87 A, 431–445.
OerlemanS, j. (1994): Quanti�ying gl�bal war�ing �r��
the retreat �� glaciers. In: Science 264, 243–245.
��i:10.1126/science.264.5156.243
– (2001): Glaciers and cli�ate change. R�tterda�.
– (2005): Extracting a cli�ate signal �r�� 169 glacier rec�rds. In: Science 308, 675 – 677. ��i:10.1126/sci- ence.1107046
– (2007): Esti�ating resp�nse ti�es �� Vadret da M�rte�M�rte- ratsch, Vadret da Palü, Briksdalsbreen and �igards�, Vadret da Palü, Briksdalsbreen and �igards- breen �r�� their length rec�rd. In: J�urnal �� Glaci�l- ogy 53, 357–362. Doi:10.3189/002214307783258387 POHjOla, V. a. and rOgerS, j. C. (1997): At��spheric
circulati�n and variati�ns in Scandinavian glacier
�ass balance. In: Quaternary Research 47, 29–36.
��i:10.1006/qres.1996.1859
raSmuSSen, l. a. and andreaSSen, l. m. (2005): Seas�nal
�ass�balance gradients in ��rway. In: J�urnal �� Glaci- ology 51, 601–606. Doi:10.3189/172756505781828990 raSmuSSen, l. a. and COnWay, H. (2001): Esti�ating
S�uth Cascade Glacier (Washingt�n, U.S.A.) �ass bal- ance �r�� a distant radi�s�nde and c��paris�n with Blue Glacier. In: J�urnal �� Glaci�l�gy 47, 579–588.
Doi:10.3189/172756501781831873
Salinger, m. j.; Heine m. j. and burrOWS, C. j. (1983): Vari- ati�ns �� St�cking (Te Wae Wae) Glacier, M�unt C��k, and cli�ate relati�nships. In: �ew Zealand J�urnal ��
Science 26, 321–338.
SOrteberg, a. and anderSen, m. S. (2008): Regional pre- cipitati�n and te�perature changes ��r ��rway 2010 and 2025 – Results �r�� 11 si�ulati�ns using 8 di��erent regi�nal ��dels. Rep�rt Series �� the Bjerknes Centre
�� Cli�ate Research 28. Bergen.
Steiner, d.; Pauling, a.; nuSSbaumer, S. u.; neSje, a.; lu-
terbaCHer, j.; Wanner, H. and ZumbüHl, H. j. (2008):
Sensitivity �� Eur�pean glaciers t� precipitati�n and te�perature – tw� case studies. In: Cli�atic Change 90, 413–441. ��i:10.1007/s10584�008�9393�1
Winkler, S. (1996): Fr�nt variati�ns �� �utlet glaciers �r��
J�stedalsbreen, western ��rway, during the twentieth century. In: ��rges Ge�l�giske Undersøkelse Bulletin 431, 33–47.
– (2002): V�n der „Kleinen Eiszeit“ zu� „gl�balen Glet�(2002): V�n der „Kleinen Eiszeit“ zu� „gl�balen Glet- scherrückzug“ � eignen sich Gletscher als Kli�azeugen?
Akade�ie der Wissenscha�ten und der Literatur Mainz.
Abhandlungen der Mathe�atisch�naturwissenscha�tli- chen Klasse 3. Mainz.
– (2008): 11 years �� bi�annual glacial length �easure�ents.
In: kjøllmOen, b. (ed.) (2008): Glaci�l�gical investiga- tion in Norway 2007. NVE Rapp�rt 2008/1. Osl�, 88.
– (2009): Gletscher und ihre Landscha�ten – eine illustrierte Ein�ührung. �ar�stadt.
Winkler, S.; elVeHøy, H. and neSje, a. (2009): Glacier fluctuati�ns �� J�stedalsbreen, western ��rway, dur- ing the past 20 years: the sensitive resp�nse �� �ari- ti�e ��untain glaciers. In: The H�l�cene 19, 395–414.
Doi:10.1177/0959683608101390
ZemP, m.; rOer, i.; kääb, a.; HOelZle, m.; Paul, f. and Haeberli, W. (2008): Gl�bal glaciers changes: �acts and figures. Zürich.
Authors Privatd�zent �r. Ste�an Winkler
�epart�ent �� Ge�graphy, University �� Würzburg, A� Hubland, 97074 Würzburg, Ger�any, ste�an.winkler@uni�wuerzburg.de
Pr��ess�r Atle �esje
�epart�ent �� Earth Sciences, University �� Bergen, Allégt. 41, 5007 Bergen
Norway, atle.nesje@ge�.uib.n�
