[ Pobierz całość w formacie PDF ]
Chapter 10
Response of River Runoff in the Cryolithic
Zone of Eastern Siberia (Lena River Basin)
to Future Climate Warming
A.G. Georgiadi, I.P. Milyukova, and E.A. Kashutina
Abstract
During the last several decades significant climate warming has been
observed in the permafrost regions of Eastern Siberia. Observed environmental
changes include increasing air temperature and to a lesser degree precipitation.
Changes in regional climate are accompanied by changes in river runoff.
Seasonal and long-term changes of river runoff in different parts of the Lena
river basin are characterized by distinct differences that can be explained by
regional distinctions of climatic conditions, types and properties of permafrost,
character of relief, hydrogeological conditions, etc. These factors determine the
non-uniform response of river runoff amount and seasonal distribution to contem-
porary climate changes within the Lena river basin. Over the past 15–20 years river
runoff has increased in different parts of the Lena river basin, but the scale of this
increase over its territory is quite variable.
Hydrological responses to climate warming have been evaluated for the plain
part of the Lena river basin based on a macroscale hydrological model featuring
simplified descriptions of processes developed at the Institute of Geography of
the Russian Academy of Sciences. Two atmosphere-ocean global circulation
models used by the IPCC (ECHAM4/OPY3 and GFDL-R30) were used to model
scenarios of future global climate. According to the results from hydrological
modelling the expected anthropogenic climate warming in the twenty-first century
can bring more significant river runoff changes in the Lena river basin compared to
the twentieth century.
Keywords
Cryolithic zone • Eastern Siberia • Scenarios of global climate warming
• Model of monthly water balance • River runoff changes
A.G. Georgiadi (
), I.P. Milyukova, and E.A. Kashutina
Institute of Geography, Russian Academy of Sciences, Staromonetny per., 29,
Moscow 119017, Russia
e-mail: galex50@gmail.com; mil-ira@list.ru; kategeo@mail.ru
H. Balzter (ed.),
Environmental Change in Siberia: Earth Observation,
Field Studies and Modelling
, Advances in Global Change Research 40,
DOI 10.1007/978-90-481-8641-9_10, © Springer Science+Business Media B.V. 2010
157
158
A.G. Georgiadi et al.
10.1 Introduction
The flow of river water forming in the cryolithic zone of Eurasia, especially within
the largest river basins (Lena, Yenisei, Ob’), has a strong impact upon regional
climate and the surrounding seas, affecting the chemical composition of the water,
formation of sea ice, and circulation of the Arctic and North Atlantic oceans.
Changes of river runoff volume and seasonal distribution from watersheds within
these basins can have considerable and in some cases critical impacts upon pro-
cesses occurring within them.
During recent decades significant global climate warming has been observed.
Climate warming has been largest in Northern Eurasia, particularly in the perma-
frost regions of Eastern Siberia (IPCC 2001a, b). Observed changes include rising
winter air temperature and overall soil temperature (Varlamov et al.
2002
), as well
as increasing winter and autumn precipitation (Chapman and Walsh
1993
; Razuvaev
et al.
1996
; Serreze et al.
2000
; Wang and Cho
1997
). These changes in regional
climate have been accompanied by changes in river runoff within Eastern Siberia
(Georgievsky et al.
1999
; Savelieva et al.
2000
; Yang et al.
2002
; Simonov and
Khristoforov
2005
; Georgiadi et al. 2008).
Climate model projections of future climate related to increasing greenhouse
gases in the atmosphere show that the permafrost regions of Eastern Siberia could
be subject to one of the most notable changes (IPCC 2001). Regional climate
warming will cause a considerable increase in soil temperature and consequent
permafrost thawing (Anisimov et al.
1997
; Demchenko et al.
2002
; Malevsky-
Malevich et al.
2001
). These changes could alter the river runoff regime, particu-
larly changes in the intra-annual distribution. Possible changes of the freshwater
and heat inflow to Arctic have potentially important implications for ocean circula-
tion and climate outside the region.
Until now, the processes regulating the water cycle in large basins have not been
studied in sufficient detail, in particular their response to climate warming. The Lena
River basin in Eastern Siberia was selected as one of the main regions of investiga-
tion. It is one of the world’s largest basins (its area is 2,488,000 km
2
, and the length
is 4,400 km) and is almost completely covered by several metres of permafrost (with
a thin seasonally-thawing active layer during the short summer). In addition, the
Lena basin is characterized by low anthropogenic influence owing to low population
density. The analysis of the Lena basin is also facilitated by a large intensive research
programs carried out by Russian, bilateral, and international teams in the last
15 years. The most significant contribution to the improvement of our understanding
of the processes in this region was achieved by field experiments that were conducted
in 1996–2004 within the framework of the international program on studying of
energy and water cycles (GEWEX) in Asia - Asian Monsoon Experiment (GAME).
We have developed a model of the Lena river basin; exploiting data collected from
GAME, and examined the changes in Lena runoff volume and seasonal distribution.
We analyze changes in runoff in response to climate model projections forced by
two scenarios of future climate, and project changes in runoff into the next century.
10 Response of River Runoff in the Cryolithic Zone of Eastern Siberia
159
10.2 Methods
10.2.1 Model of Monthly Water Balance
The initial versions of the model and its application to the largest river basins of the
Russian plain are considered in detail in Georgiadi and Milyukova (
2002, 2006
). In
the model the basic processes of the hydrological cycle are described: infiltration
and moisture accumulation in the soil, evaporation (on the basis of the modified
Thornthwhaite’s method (Willmott et al.
1985
), accumulation of water in the snow
pack and snow melting on the basis of Komarov’s method (Anon
1989
), formation
of surface, subsurface and groundwater flows in the rivers and full river runoff. The
model can take into account macroscale heterogeneity of hydrometeorological
fields and other territory characteristics, allowing a degree of reliability in the mod-
elling of the river runoff changes. In the model of the monthly water balance the
changes of the river runoff and other water balance elements are estimated in units
of a regular grid, which facilitates the coupling of the model with climate model
simulations.
In adapting the model for the conditions of permafrost soils it was assumed that
the process of runoff formation in general can be presented as follows. In the cold
season, precipitation arrives in solid form (snow) and accumulates. The top layers
of soil at negative air temperatures freeze through, forming a practically water-
proof layer. Positive air temperature signals the start of snow melt and soil thaw.
Precipitation and melt water, getting to the surface of the catchment, partly evapo-
rate from the surface, and the remainder seeps into the thawed active layer of the
soil, the thickness of which gradually increases as the soil thaw progresses. At the
same time, part of the water filters into the underlying horizons of underground
water. If the soil layer becomes saturated to up to field capacity, a fast flow (mainly
surface flow and flow from active layer of the supra-permafrost zone) is formed
from the surplus moisture. The moisture held in underground water-bearing hori-
zons is redistributed, with part forming surface flow and part infiltrating into deeper
water-bearing horizons (i.e. groundwater) outside a zone of active water
exchange.
The model is based on a conservation equation of average long-term monthly
water balance of river catchments. In general it can be written down for each cell
of a regular grid in the following way:
Q t Q t P t E t I t dW dt
() () () () () /
+ =−−−
d
(10.1)
where
Q
s
(
t
) is the total surface and subsurface (seasonal active layer) flow (mm);
Q
gr
(
t
) is groundwater flow (mm). The sum of
Q
s
(
t
) and
Q
gr
(
t
) is the full river runoff.
P
(
t
)is atmospheric precipitation (mm);
E
(
t
)is evaporation (mm);
I
d
(
t
)is infiltra-
tion of water to deep horizons of underground water outside of the active water
exchange zone (mm),
dW/dt
is the change of the water amount in the active
water exchange zone of the river basin,
t
is time.
s
gr
160
A.G. Georgiadi et al.
For the water balance calculations in the Lena river basin a number of model
blocks has been modified for the conditions of perennially frozen ground, and
also blocks for calculations of seasonal soil thawing have been set up. In this version
of the model two fundamentally different procedures are used: one allows to esti-
mate the dynamics of seasonal thawing and freezing of the active layer and the
other is used for areas where there is no perennially frozen ground, allowing to
calculate the dynamics of the seasonally frozen layer depth and its influence on the
flow formation.
For the calculation of the progression of the thawing (freezing) front, Eq. 10.2
is used (Pavlov
1979
; Belchikov and Koren
1979
), which has been established as a
simplified solution of the classical single-front Stefan problem. The Stefan problem
is a one-dimensional ordinary differential equation of heat conduction with a liquid/
solid phase change at the boundary. In permafrost, the problem becomes a two-
front problem since freezing does not only progress from above but also from
below due to a store of cold in the deep frozen layer. The equation was defined
assuming that the inflow of cold (heat) from below through the border of thawing
(freezing) is constant (or a negligibly small quantity in comparison with the heat
(cold) flux from the atmosphere). In this equation the height and density of snow
cover are taken into account, and the initial temperature of soil thawing (freezing)
may not be equal to 0°C.
{
2
( )
}
12
/
Z St
=± + + ± λ ∆ ρ
St Z
j j1
−
2
T t / LW
j
j
(10.2)
p
B
Z
where
Z
j
is the depth of the thawing front at day
j
and
St
j
is a variable which
takes into account the availability and depth of the snow cover when calculating soil
freezing or thawing. Note that when performing calculations for the changing
boundary of soil freezing,
( )
St
j
ll
H
, but for the changing boundary of soil
j
c
St A
.
L
is the specific heat of ice fusion, l
c
is the thermal conduc-
tivity of snow, l is the thermal conductivity of frozen (at freezing) and thawed
(at thawing) soil,
H
is the height of the snow layer, and T
p
is the temperature of
underlying ground, A
k
is a parameter; W
z
is volumetric soil moisture at the front of
thawing (as a fraction of 1),
r
B
is the density of water, ∆
t
is the time interval (days).
In the absence of a snow cover the dependence between air temperature and
ground temperature was fixed, with the factor of permafrost being taken into
account. Based on observation data, it is possible to set
= ;
j
= /
k
T Tk
(
k
t
= 1.0 » 1.3),
where
T
is daily average air temperature and
T
p
is daily average soil temperature.
The thermal conductivity of thawed and frozen soil grounds was calculated by
the relation
j
j
p
t
l
= + −−
( 0.1 1.1) 0.1
p sk
(10.3)
where
sk
j
is the relative density of the soil particle size fraction (g/cm
3
),
k
= 1.3−1.4 for loamy
sand and clay (the first value for thawing soil, the second – for frozing soils),
W
is
the volumetric soil moisture in %. The thermal conductivity of snow was calculated
k
= 1.4−1.5 for sandy loam,
p
j j
= /
thawing,
l
kj W W
k
= 1.5−1.7 for sand,
p
p
10 Response of River Runoff in the Cryolithic Zone of Eastern Siberia
161
as
λ
= + +
4.32 1.64 518.4
r
r
4
, where
r
c
is the snow density changing over time:
0
c
c
c
r r rr rr r r r
=
j
+
j
k
,
j
, =
j
j-1
+
k
,
is the initial snow density,
K
c
is a empirical
c
c
c
cc
c c
c
c
0
parameter.
Infiltration to the underground water horizon of active water exchange (
I
) for
day
j
was defined by the following relations:
in the absence of permafrost:
I kP M E WW
j
= +− /
(
j
j
j
)( )
j
(10.4)
p
f
fc
in the presence of permafrost:
I tk P M E WW
j
= +−
(
j
j
j
)( )
j
/
(10.5)
p
f f
fc
where
M
is daily snowmelt (mm),
W
is the soil moisture reserve at field capacity
fc
t
is a parameter depen-
dent on the type of permafrost distribution in the given cell. Other symbols are
given above.
The ground water runoff (mm of depth) was calculated as
t
is a factor of vertical infiltration in ground water,
f
f
Q kHH
j
=
(
j
/
)
(10.6)
gr
gr gr
max
where
H
max
is the maximum capacity of the aquifer,
k
is a factor of ground water
gr
runoff, and
H
is the capacity of the aquifer on day
j
, calculated as
gr
HH I Q I
j
= +− −
j
−
1
j
j
−
1
j
(10.7)
gr
gr
p
gr
d
The moisture infiltration
I
d
into the deep horizons of groundwater outside of the
active water exchange zone was set as a constant. The change of the groundwater
capacity of active water exchange for day
j
was calculated by:
∆
HHH
j
=−
1
j
j
−
(10.8)
gr
gr
gr
Filtration of water into the aquifer of active water exchange (
I
) and groundwater
Q
) for different types of permafrost soil (continuous, discontinuous, sporadic
and absent) were calculated using empirical factors derived from field observations.
10.2.2 General Scheme of Calculations
To initialise the model we used monthly average long-term data of air temperature
and precipitation, soil moisture and physical soil characteristics (field capacity, soil
volume and bulk density), types of permafrost lateral continuity (continuous, dis-
continuous, sporadic, absence of permafrost), the maximum depth of the active
layer, and runoff for each grid cell.
p
(mm),
p
flow (
gr
[ Pobierz całość w formacie PDF ]