; from which
extensive data streams enter IMO and are utilized
for forecasts and research purposes.
Dissemination
The main dissemination of IMO is in the form of
forecasts and warnings; through radio, T V, direct di-
alog with stakeholders and through IMO‘s web-site
(vedur.is). Additionally, the web provides compre-
hensive real-time data on the weather, earthquakes
and deformation, as well
/media/vedurstofan/utgafa/arsskyrslur/VED_AnnualReport-2013_screen.pdf
) and Jónsdóttir (2008).
Therefore, comparison of measured and simulated water balance cannot be di-
rectly used for validation of the model-generated precipitation. According to the
non-scaled MM5 output for the period 1961–1990, mean precipitation for the
whole of Iceland was 1790 mm y−1. After scaling the precipitation, this value
was reduced to 1750mm y−1, i.e. by approximately 2%. This difference
/media/ces/Paper-Olafur-Rognvaldsson_92.pdf
are largest. Positive sensible heat fluxes also occur over
the interior regions of Vatnajökull and Hofsjökull at around noon on 27 July, due to the cold
northeasterly flow over the glaciers (see Figure 10). However, along the edges and on the other
icecaps, sensible heat fluxes under clear skies are directed from the atmosphere to the snow. On 3
August, with clear skies, weak winds, and with above freezing
/media/vedurstofan/utgafa/skyrslur/2015/VI_2015_006.pdf
the
interior of the ice sheet is somewhat too dry (Fig. 9b). By average a mean negative precipitation
bias of 0.16myr−1 results which equals 43% of the mean from Burgess et al. (2010) (Table 2).
6 Bias Correction and Future Scenario Runs
After having specified a number of biases in the RCM output the model runs were repeated
with bias-corrected RCM data.
To correct the temporal bias of Ta, daily
/media/ces/ces_geus_paakitsoq_full_report.pdf
m
J
M5 [C°] -3
obs. [C°] -4
nce 1
re 5. Comp
26); an int
temperatu
this system
y gridded v
picion abo
-Jökulsá w
similar dif
han observ
h elevation
ces the effe
months No
ly only on
high the tem
n band wi
refore be s
onthly tem
an Feb Ma
.2 -3.1 -3.
.3 -4.1 -3.
.1 1.0 0.6
arison of m
erpolation
re is shown
atic differe
alues, see T
ut the qual
atershed; b
ference wa
ations for t
gradient fo
/media/ces/2010_017.pdf
at each site i with the same method used to estimate qR(D;T ), but instead of pooling
AMF series for a given duration D from different sites, the estimation is made individually for
each site i by pooling AMF series for different durations D (see Crochet, 2012c). The index
flood µi(D), is modeled at each site i as a continuous function of D, as follows:
µi(D) =
µi
1+(D=Di)li
; (5)
where µi, Di/media/vedurstofan/utgafa/skyrslur/2014/VI_2014_001.pdf
Veðurstofa Íslands
2 Almannavarnadeild Ríkislögreglustjóra
3 Jarðvísindastofnun Háskólans
4 Istituto Nazionale di Geofisica e Vulcanologia (INGV), Bologna
5 Istituto Nazionale di Geofisica e Vulcanologia (INGV), Pisa
6 Jarðvísindadeild Háskóla Íslands
7 Agricultural University of Iceland
8 Consultant
Skýrsla nr. Dags. ISSN Opin Lokuð
VÍ 2020-011 Desember 2020 1670-8261 Skilmálar:
Heiti skýrslu
/media/vedurstofan-utgafa-2020/VI_2020_011_en.pdf
The hydrological simulations were performed with the Wa-
tershed Simulation and Forecasting System (WSFS) developed
and operated in the Finnish Environment Institute (Vehviläinen
et al., 2005). The WSFS is used in Finland for operational hydrolog-
ical forecasting and flood warnings (www.environment.fi/water-
forecast/), regulation planning and research purposes
(Vehviläinen and Huttunen, 1997
/media/ces/Journal_of_Hydrology_Veijalainen_etal.pdf