Geoscience Reference
In-Depth Information
Furthermore, an ensuing question that relates to all of the above is: What critical gaps
and key limitations exist in each of these topics, and is there a basis to rationally prioritize
how to address those limitations by hydrological monitoring development?
We here synthesize information on these topics by integrating a set of previous studies
of Arctic hydro-climatic change (Bring and Destouni
2009
,
2011
,
2013
; Dyurgerov et al.
2010
). We also extend the results from these reports with an additional analysis of GCM
projection performance for major Arctic hydrological basins.
2 Methods Summary
In the following, we summarize the data and methods used in the synthesized previous studies
(Bring and Destouni
2009
,
2011
,
2013
; Dyurgerov et al.
2010
) and the novel extensions in this
paper. For a more detailed account of the methods used in the previous studies, we refer to the
aforementioned publications. We have in this review re-evaluated some of the previous results
by correcting observed precipitation data sets for gauge undercatch and orographic effects. To
this end, we used two global data sets by Adam and Lettenmaier (
2003
) and Adam et al. (
2006
).
However, as precipitation corrections in other cases have been known to result in overestimated
precipitation (M¨chel et al.
2012
), we here treat the originally reported observations and the
fully corrected values as two ends of a range. As we have no knowledge of the probability
distribution of values within this range, we treat the middle of the range (i.e., the average of the
original and the corrected values) as the best estimate of observed precipitation.
In general, the hydrological drainage basin has been the fundamental basis for the
investigations in this study. The drainage basin constitutes a physically consistent
boundary for closing the flow balance of water and the mass balances of constituents
transported by water, which makes it a relevant scale for addressing both science and
management problems.
In the paper, we focus primarily on the 14 largest watersheds in the PADB. Each of
them is at least 200,000 km
2
in size, and their combined drainage covers 13.8 million km
2
(green and blue basins in the map of the study area in Fig.
1
). These 14 basins are
sufficiently large to allow an analysis of GCM data. Figure
2
shows the characteristic
temperature and precipitation conditions, and recent observed changes for those parame-
ters, in the 14 major basins.
The analyses in the surveyed and synthesized publications concern slightly varying time
periods and, in general, involve comparisons of time periods of different lengths. Although it
would be desirable to always compare the same time periods, the differences between the
surveyed publications are relatively small in this regard. Several publications also study the
longest possible period since 1990 for which there are data available, and compare that period
with the reference of 1961-1990. Since the recent period from 1991 to (near-) present time is
considerably shorter than the 30 years of the climatological reference period 1961-1990, we
here regard the changes from the latter to the recent period as deviations from the reference-
time climate and not necessarily as climate change. With inter-annual variability of various
magnitudes for different parameters and basins, the future 30-year climate starting from 1991
may differ from these shorter-term deviations.
2.1 GCM Projections Across Arctic Basins
We first examine two successive generations of GCMs, which form the basis for the two
latest available full IPCC assessment reports, the TAR and AR4. We here compare the
