Title: Variation of snow accumulation and annual river runoff in North Eurasia and their relation to atmosp
1Variation of snow accumulation and annual river
runoff in North Eurasia and their relation to
atmospheric circulation changes in the 20th
century _________________________________________
________
Valeria Popova Institute of geography, Russian
Academy of sciences
- Supported by the INTAS grant 00-77
2Surface air temperature anomalies in winter, Co,
averaged for the period 1989-2001, compared with
the mean for 1951-1980Shmakin and Popova, 2004
Recent warming over the Northern Eurasia
3Differences of surface pressure anomalies
between the years with positive (INAOgt1) and
negative (INAOlt-1 ) phases of NAO for the 100
year period of observations
Variations of winter air surface temperature
anomalies, averaged over the FSU territory, and
winter NAO index
4Surface air temperature anomalies, associated
with North Atlantic Oscillation (NAO) changes
INAO gt 1.0 (n7)
INAO lt - 1.0 (n4)
Pattern, standardized winter anomalies of INAO ,
in 1951-2002 (left panel) and average January air
temperature anomalies (oC) for anomalous INAO
(right panel). Painted areas indicate
statistically significant anomalies (plt0.05) n
number of cases.
5Polar Eurasian (POL)
IPOL gt 1.0 (n4)
IPOL lt - 1.0 (n8)
Pattern, standardized winter anomalies of IPOL,
in 1951-2002 (left panel) and average January air
temperature anomalies (oC) for anomalous IPOL
(right panel). Painted areas indicate
statistically significant anomalies (plt0.05) n
number of cases.
6Pacific - North American (PNA)
IPNA gt 1.0 (n8)
IPNA lt - 1.0 (n4)
Pattern, standardized winter anomalies of IPNA,
in 1951-2002 (left panel) and average January air
temperature anomalies (oC) for anomalous IPNA
(right panel). Painted areas indicate
statistically significant anomalies (plt0.05) n
number of cases.
7 Scandinavian (SCAND)
ISCAND gt 0.8 (n3)
ISCAND lt - 0.8 (n4)
Pattern, standardized winter anomalies of ISCAND,
in 1951-2002 (left panel) and average January air
temperature anomalies (oC) for anomalous ISCAND
(right panel). Painted areas indicate
statistically significant anomalies (plt0.05) n
number of cases.
8Objectives
- to reveal the regions with homogeneous
fluctuations of snow depth - to assess impact of atmospheric circulation
patterns, dominant in Northern Hemisphere, on
snow depth variability - to evaluate impact of interannual snow depth
variations on changes of annual runoff from
large-scale river basins - to detect recent climate change signal in winter
snow accumulation and river runoff
9Data
- Average February snow depth, derived from the
daily data collected by first order
meteorological station from 1936-2001 283
stations - former USSR territory (Russia
Institute of Hydrometeorological Information
RIHMI), Finland -5 stations, and Norway 2
stations. - Northern Hemisphere teleconnection patterns for
the 700 hPa height Barnston and Livezey,1987
North Atlantic Oscillation (NAO, differs from
NAO, evaluated based on sea level pressure data),
Polar-Eurasia (Pol), Pacific-North American
(PNA), West Pacific (WP), Scandinavian (Scand)
and standardized anomalies of those
teleconnection indices INAO, IPOL, IWP, IPNA,
ISCAND (1950-2002), derived as a result of
rotated principal component analysis (PCA)
http//www.cpc.ncep.noaa.gov/data/teledoc - Index NAO(INAO ) - 1879-2001
http//www.cru.uea.ac.uk/ftpdata/nao.dat
10Data
- Annual runoff for three major Siberian rivers Ob
(Salekhard), Yenisey (Igarka), Lena (Kusur) was
calculated on the base on monthly discharge data
from the beginning of observation to 1999 Arctic
river runoff. http//www.R-arcticnet.unh.edu.
Volga annual runoff time-series accounts for 112
years (1879-1999) - Areas of the basins are determined with 0.5o
resolution Oki T., Sud Y. Design of total runoff
integrating pathways (TRIP) a global river
channel network. Earth Interactions, 1998,
vol.2, N 11 -
11Methods
- Original data for each year were interpolated
into grid points with 5X5o resolution, 161 grid
point over the area covered 40-70oN, 20-160oE
(after excluding areas containing no data) is
selected to EOF analysis. Statistically
significant PCs were retained for Varimax
rotation and further consideration. - Time series of rotated PCs were correlated to the
circulation indices to reveal their effect on
snow depth variation. Impact of Northern
Hemisphere circulation modes was evaluated using
forward stepwise multiply regression.
12Methods
- Snow depth time-series for the grid points were
used to derive patterns of of the following three
parameters mean snow depth difference between
average snow depth anomalies for the years with
INAOgt1 and INAOlt-1 correlation coefficient
between snow depth and annual runoff. - For the long Volga runoff time-series, variance
spectrum and coherency function with winter NAO
variations were estimated. Cross-correlation and
auto-correlation functions were estimated, as
well.
13Correlation Coefficients between snow-depth PCs
and winter circulation Indices (1951-2001)
Marked coefficients are significant at p lt0,05
14Pattern (upper panel) and time series (lower
panel) of February snow depth PC1 1 PC1,
based on data of observation 2
calculated PC1, basing on stepwise
multiple regression with included
circulation indices INAO , ISCAND ,
IPNA , IPOL
15Multiple regression on snow depth PC1 and the
fractions of the variance in snow depth (R2)
associated with circulation indices for two
periods. Numbers in bold are statistically
significant (plt0.05)
16Multiple regression on snow depth PC1 and the
fractions of the variance in surface air
temperature over the Northern Eurasia (R2)
associated with circulation indices for two
periods. Numbers in bold are statistically
significant (plt0.05)
17Pattern (upper panel) and time series (lower
panel) of February snow depth PC2 1 PC2,
based on data of observation2
calculated PC2, basing on stepwise
multiple regression with included
circulation indices INAO , IPOL
18Pattern (upper panel) and time series (lower
panel) of February snow depth PC3 1 PC3,
based on data of observation2
calculated PC3, basing on stepwise
multiple regression with included
circulation index IWP
19Pattern (upper panel) and time series (lower
panel) of February snow depth PC4 1 PC4,
based on data of observation2
calculated PC4, basing on stepwise
multiple regression with included
circulation index ISCAND
20Pattern (upper panel) and time series (lower
panel) of February snow depth PC5 1 PC5,
based on data of observation2
calculated PC5, basing on stepwise
multiple regression with included
circulation index IWP
21Location of the river basins Volga, Ob, Yenisey,
Lena (painted areas), and limits of the regions,
homogeneous with respect to variations of winter
snow accumulation (dashed lines)
22Annual cycle of runoff, km3/month
Ob
Volga
Lena
Yenisey
23Interannual variations of annual Yenisey runoff
snow depth averaged over the basin and winter NAO
index (annual and 5-year smoothed)
24Correlation between annual Yenisey runoff and
winter snow accumulation (1936-2001)
Distribution of the mean snow depth, cm (a), the
difference between average snow depth for the
years with INAOgt1 and INAOlt-1, cm (b), and
correlation coefficient between snow depth and
annual runoff (c) cross-correlation function of
annual Yenisey runoff and average snow depth (d).
a
b
d
c
25Interannual variations of snow depth averaged
over the basin, annual Volga runoff and winter
NAO index (annual and 5-year smoothed)
26Correlation between annual Volga runoff
variability and winter NAO index, INAO,
(1879-1991).Variance spectrum of Volga annual
runoff (a), winter average INAO (b) and coherency
function (c). Vertical (a, b) and horizontal (c)
lines indicate 90 and 95 confidence level,
respectively.
b
27Correlation between annual Volga runoff and
winter snow accumulation (1936-2001)
Distribution of the mean snow depth, cm (a),
the difference between average snow depth for the
years with INAOgt1 and INAOlt-1, cm (b), and
correlation coefficient between snow depth and
annual runoff (c) auto-correlation function of
annual Volga runoff (d).
a
b
d
c
28Interannual variations of annual Ob runoff, snow
depth averaged over the basin and winter NAO
index (annual and 5-year smoothed).
29Correlation between annual Ob runoff and winter
snow accumulation (1936-2001)Distribution of
the mean snow depth, cm (a), the difference
between average snow depth for the years with
INAOgt1 and INAOlt-1, cm (b), and correlation
coefficient between snow depth and annual runoff
(c) auto-correlation function of annual Ob
runoff (d)
a
b
d
c
30Interannual variations of annual Lena runoff,
snow depth averaged over the basin and winter NAO
index (annual and 5-year smoothed)
31Correlation between annual Lena runoff and winter
snow accumulation (1936-2001)Distribution of
the mean snow depth, cm (a), the difference
between average snow depth for the years with
INAOgt1 and INAOlt-1, cm (b), and correlation
coefficient between snow depth and annual runoff
(c) cross-correlation function of annual Lena
runoff and average snow depth (d)
a
b
d
c
32 Conclusions
- Winter snow accumulation over the Northern
Eurasia is controlled by certain Northern
Hemisphere atmospheric circulation modes. - Regions homogeneous with respect to interannual
variations of snow depth are revealed. Temporal
variability of snow accumulation in the regions
differs by the share of low- frequency variations
as well as the impact of certain circulation
indices.
33 Conclusions
- Increase of snow accumulation over the vast
region, between the White Sea and Yakutia, can be
considered as one of the most important indicator
of the recent global warming. Positive phase of
NAO and enhanced westerlies appear to be a major
circulation factor both of temperature and snow
accumulation positive trend since 1970s. - For Volga and Yenisey basins the positive trend
of runoff since 1970s is caused by increased snow
accumulation associated with positive phase of
NAO. Though decadal variations of snow depth over
all of the studied basins are reflected in the
annual runoff, in Ob and Lena basins their
long-term changes do not coincide due to
prevailing share of summer precipitation
variations.