Jökull - 01.01.2013, Page 80

P. Crochet
Table 4. Differences in the median or mean of various hydro-climate variables (Table 2) between warm and
cold years (∆ = Median[warm] – Median[cold] and θ = E[warm]/E[cold]). Statistically significant changes at
the 10% and 5% levels are marked with a * and **, respectively. – Mismunur á milli hlýrra og kaldra ára (∆ =
hlýtt – kalt, θ = hlýtt/kalt). Tölfræðilegur marktækur mismunur miðað við 10% líkur er sýndur með * og miðað
við 5% líkur með **.
Variable / Gauging station VHM- VHM- VHM- VHM- VHM- VHM- VHM- VHM-
19 10 26 145 144 66 64 96
∆ (AT) (◦C) 1.4** 1.3** 1.2** 1.1** 1.1** 1.2** 1.3** 1.3**
∆ (AP)/AP-cold (%) 6 2 –4 –12 –17** 18. 16* 21**
∆ (AR)/AR-cold (%) 53** 12 –3 15 6 28** 27** 50**
∆ (AWR)/AWR-cold (%) 6* 2 –2 2 –2 22** 18** 20**
∆ (AQ)/AQ-cold (%) 13** –6 –11 9 4 18** 11** 12**
∆ (AMS)/AMS-cold (%) –25 –41* –40** –36** –39** –8 –20 –16**
∆ (DAMS) (days) –20** –41 –71* –20** –21** –26** –11 –16
∆ (ASD) (days) –63* –38** –60** –18* –29** –33 –45* –34*
∆ (ASR)/ASR-cold (%) –16 1 –6 –21** –26** –1 –7 –5*
∆ (AGR)/AGR-cold (%) NA NA NA 24** 29** 39** 43** 32**
∆ (AGD) (days) NA NA NA 21** 20** 24** 22** 17**
∆ (CTS) (days) –30** –30** –67** –37** –27** –33** –42** –28**
∆ (CTW) (days) –46** –40** –49** –25** –20** –24** –43** –30**
∆ (CTQ) (days) –64** –35** –48.5** –27** –15** –7** –15** –35**
∆ (AMF)/AMF-cold (%) 27 –21 –29** 1 –15 8** 16 65**
∆ (DAMF) (days) –186 –30** –77* –25** –25** –25 2 –52**
∆ (SMF)/SMF-cold (%) –21 –24 –25** –9 –15 33 13 39*
∆ (DSMF) (days) 0 –23** –31 –25** –25** –37** –43** –49**
θ (FOR): POT-3 1.5 1.2 0.9 1.2 0.7 1.7* 1.4 3.3**
θ (FOR): POT-2 1.2 1.2 0.9 0.9 0.7 2.9** 2.4** 4.1**
with elevated temperatures in catchments where total
water input runoff (AWR) increased, in the south, west
and northwest, but remained unchanged elsewhere.
Higher temperatures significantly shifted the cen-
ter of volume dates of snowmelt (CTS), total water
input (CTW) and streamflow (CTQ) of all catchments,
by several weeks in all three cases. This indicates that
the earlier timing of the bulk of snowmelt runoff con-
trols the timing of the bulk of streamflow of the rivers
in question. An increased ratio of winter rainfall to
snowfall could also account for some of these stream-
flow timing changes for some catchments. A signif-
icant correlation was usually observed between CTQ
and CTS (0.2 ≤ R2 ≤ 0.68; p <1%) and between CTQ
and CTW (0.16 ≤ R2 ≤ 0.64; p <1%). The smallest
CTQ shifts were observed at VHM-64 and VHM-66,
probably due to the buffering effect of groundwater
aquifers. The timing of the spring flow peak (DSMF)
was shifted several weeks earlier in warm years com-
pared to cold years in most catchments, due to an early
onset of snowmelt, except for the two northernmost
catchments. However, the spring flow peak magnitude
remained mostly the same between cold and warm
years.
Annual flow peaks are often observed in spring,
during the snowmelt season, or autumn and winter,
during the passage of cyclones with heavy rain and/or
snowmelt. The proportion of these different types of
events is variable between watersheds. The timing
of the annual flow peak was earlier in warm years
than in cold years in the north and south, similar
as for the spring flow peak, and also in the north-
east, but remained unchanged for the northwestern
catchment and the two southwestern catchments with
large groundwater aquifers. The annual flow peak
magnitude significantly increased in response to in-
80 JÖKULL No. 63, 2013