Transcript Annual difference from present in evaporation mines

Atmospherically – induced hazards in the
coastal zone and the possibility of their
decadal and centennial prediction
A.Kislov, G.Surkova, D.Gushina, P.Toropov,
D.Blinov.
Dpt. Of Meteorology and Climatology
NRAL, 14.12.2012
Contents
1.
2.
3.
4.
5.
Preparation of the meteorological
observations for all working groups
Preparation of initial conditions for the start
of the "sea models”
Hazardous weather explicit modeling (bora,
cyclones) in the COSMO and WRF
Identification of synoptic conditions,
corresponding to different hazards
Projection of extreme atmospheric processes
under the climate change
Contents
1.
2.
3.
4.
5.
Preparation of the meteorological
observations for all working groups
Preparation of initial conditions for the start of the "sea models”
Hazardous weather explicit modeling (bora, cyclones) in the COSMO and WRF
Identification of synoptic conditions, corresponding to different hazards
Projection of extreme atmospheric processes under the climate change
Archive of station data
Variables:
OBSERVATION PERIOD:
Pressure, surface air
maximum period of the data span is 01.01.1871 to 01.01.2001
temperature, water
most stations (more 60%) operated during the period 1936-1990
vapor
pressure,
The measurements were made:
wind speed and
3times a day before 1936
direction,
4 times a day at mean astronomic times during the 1936-1965
characteristics
of
period.
cloudiness, current
8 times a day at United Time Coordinate (UTC) since 01.01. 1966
weather
structure of the data archive:
Archive contains three types of data files (ID –
index of station):
STN_ID.dat (2095 files, 1 per station), data in
ASCII
STN_ID.flg (2095 files, 1 per station), quality
marks of data
ussr_hly_stn.list.txt (1 file with essential
station metadata such as station identifier,
coordinates, elevation, date of the first and the
last records, and station name)
Density measuring network
Calendars of hazardous weather
Hazardous weather (HW) cases selection – three calendars
1) Calendar for extreme observed wave storms and surges (19482012)
2) Calendar for wave storms with significant wave height 4 m
modelled by wave model SWAN (1948-2012)
3) Calendar of hazardous weather (HW) in the future (2046-2065) –
detected by the statistical methods, data of numerical climate
simulations (CMIP3)
Contents
1.
2.
3.
4.
5.
Preparation of the meteorological observations for all working groups
Preparation of initial conditions for the
start of the "sea models”
Hazards explicit modeling (bora, cyclones) in the COSMO and WRF
Identification of synoptic conditions, corresponding to different hazards
Projection of extreme atmospheric processes under the climate change
Initial and boundary conditions for
wave modelling
Reanalysis
Every 6 hours
Surface
WIND DATA
NCEP/NCAR (1948-2012)
1,9x1,9 degree
Mesoscale
nonhydrostatic
atmospheric
model
COSMO-RU
ERA-40 (1958-2002)
2,5x2,5 degree
ERA-Interim (19792012) , 1x1 degree
Downloading data => decoding => preparing for selected area
Contents
1.
2.
3.
4.
5.
Preparation of the meteorological observations for all working groups
Preparation of initial conditions for the start of the "sea models”
Hazardous weather explicit modeling
(bora, cyclones) in the COSMO and
WRF
Identification of synoptic conditions, corresponding to different hazards
Projection of extreme atmospheric processes under the climate change
Some results of the statistical
evaluation of forecast accuracy
a) the average forecast (blue
line) and observed (red line)
wind speed; b) the empirical
pdf of forecast (red bars) and
actual (blue bars) wind speeds
over the “test area” (along the
horizontal axis – the intervals in
m/s), c) modal values of wind
speed: forecast (red line) and
the actual value (blue line)
(along the horizontal axis –
hours forecast).
Wind speed (a,c) and wave heights (b,d)
at the time of Novorosiysk bora averaged
over 26. 01.2012
wave heights calculated
by the model SWAN
analysis of
NCEP/NCAR
1×1
WRF-ARW
On the way to the forecast of water flood in
Sochi-Tuapse in October 2010 by COSMO-RU
Precipitation during last 12 hours. Forecast for 10:00 MSK 16.10.2010
24 hours ahead
48 hours ahead
66 hours ahead
Forecast of water surface runoff by
COSMO-RU.
Runoff during previous 24 hours
Model is capable to simulate
the extreme runoff during the
flood.
Contents
1.
2.
3.
4.
5.
Preparation of the meteorological observations for all working groups
Preparation of initial conditions for the start of the "sea models”
Hazardous weather explicit modeling (bora, cyclones) in the COSMO and WRF
Identification of synoptic conditions,
corresponding to different hazards
Projection of extreme atmospheric processes under the climate change
Synoptic situations associated to the
Storm surges
various types of flood
Predictors:
trajectories of
depressions,
wind speed and
wind direction,
duration of wind
forcing
Neva River 28.10.2006 12 UTC
Don River 28.02.2005 12 UTC
Water-flow
Predictors: the main factor is
abundant precipitation. No unified
scheme of synoptic situation, but
the intensive frontal zone is
always presented
Ice-jam
Predictors: large zonal frontal zone
expanding in north-south direction,
temperature jumps precipitations fall
conditions, wind direction in the mouth
of river etc
Mzymta
26.10.1997
00 UTC
Pechora 2.06.2008 12 UTC
Contents
1.
2.
3.
4.
5.
Preparation of the meteorological observations for all working groups
Preparation of initial conditions for the start of the "sea models”
Hazardous weather explicit modeling (bora, cyclones) in the COSMO and WRF
Identification of synoptic conditions, corresponding to different hazards
Projection of extreme atmospheric
processes under the climate change
Interannual, decadal and centennial
hazardous weather projections. How
they can be predicted?
Sahel
NAO
G.Meehl
PDO
ЕNSO
CMIP5 (Coupled
Model
Intercomparison
Project)
Method of hazardous weather
projection for a long time:
step by step
Wind speed more than 15 m/s:
observation, hindcasting, projecting
The Black Sea costal
wind observations
(1948-2011)
Climate change and storm events
frequency
-time series of V  15 m/s
don’t show obvious trends;
-synoptic features for storm
events: it is reveled that
prevailing of 1st type of SLP
fields for storms took place
for the last 60 years and it is
expected to continue in the
21 century;
-climate projection based on
ECHAM5 simulation shows
slight redistribution of
monthly frequency of strong
winds and conservation of the
ration of storms SLP fields
types
The same projecting extreme wind
speed for the Caspian Sea
Caspian Sea – 137 cases of
HW (wave height >4 m)
I type
65 %
MPI-ECHAM5
II type
35 %
Change of occurrence of water flows predictor
under warmer climate in the Black sea area
• The probability of occurrence of predictors for water flows in the Black
sea region was estimated for modern climate and global warming
conditions using the outputs of ECHAM5/MPI-OM model.
• It is shown that the occurrence of intensive frontal zone in the South
of Russia will increase (decrease) in summer (winter) under warmer
climate conditions which may contribute to the increase of water
flow risks in summer.
Number of
cases
Winter
1961-1980
1981-2000
2046-2065
303
381
217
13
17
13
20
38
75
1
1
2
Summer
grad T> 18oC/1000 km
grad T>18oC/1000 km
precipitation>20 mm/day
grad T> 12oC/1000 km
grad T>12oC/1000 km
precipitation>20 mm/day
Change of intensive frontal zone occurrence in the
Black sea area
Number of cases with intensive frontal zone
Summer
Number of cases
Number of cases
Number of cases with intensive frontal zone
Winter
Years
Years
19611980
1980-2000
2046-2065
Mean
22
26
19
Rms
8.996
10.658
9.760
80.937
113.589
95.263
11
10
4
48
44
42
Dispersion
Minimum
Maximum
1961-1980
19802000
20462065
Mean
5.5
6
12
Rms
6.116
6.504
7.177
37.411
42.305
51.503
Minimum
0
0
1
Maximum
18
27
26
Dispersion
Conclusions and future plan
-efficiency of prognostic methodology for changes of hazardous
weather have been demonstrated
-model MPI-ECHAM5 shows good agreement with assessment
of frequency of storms SLP fields and used for projection of
storms frequency in 21 century; it allows to hope that this
technique can give relevant practice information
-All CMIP5 models will be used for realization of this task, based
both on different RCPs and different decadal forecasting
Thank you!
Black Sea
Frequency of each of 19 HW and climate
projection
1948-2010
reanalysis
MPI-ECHAM5
HW types for data series 2 and 3 (SWAN calendar, 1948-2010)
Black Sea – 137 cases of HW (wave height >4 m)
EOF
Surface pressure - centroids
I type
43 %
HW type
I
II
1
0,44
0,45
2
0,20
0,21
3
0,10
0,10
4
0,07
0,07
5
0,04
0,04
6
0,03
0,03
7
0,03
0,02
8
0,02
0,02
II type
57 %
Weather types are revealed by EOF
and cluster analysis – two main
statistically significant types
How we can predict unmodelable processes based on climate
simulation: application to the problem of predicting of hazardous
storm wind speeds on the shores of the Black Sea
•First, we have to establish a relationship in quantitative terms (based on
observations) between storm wind speed and sea-level pressure (SLP) field
•Second, we will test how well climate model reproduce the desired features of SLP.
•Third, to determine what changes occur in the SLP climate forecast.
•Fourth, we have to make the transition to the prediction of the studied hazard
/First & Second/
storm.
Storms over the Black Sea area
have been studied for the last 60
years based on reanalysis data
and coastal observations. A wind
speed of 15 m/s is chosen as a
threshold to detect the storm
situation. EOF analyses of SLP are
applied to identify the main types
of atmospheric circulations
causing severe winds and storm
waves. The first three EOFs cover
more than 70% of the total
dispersion in all cases. This fact
allows to create a ‘data bank’ of
filtered SLP pattern for previous
wind
storms and to compare any single
Important changes of the storm activity on
the shores of the Black Sea will not be
expected under the future global warming
scenario
/Third &Fourth /
Data source: CMIP3
Data type: daily sea
level pressure (SLP)
Climate model:
MPI-ECHAM5 (Max
Planck Institute for
Meteorology,Hamburg,
Germany)
Numerical experiments
ID:
- 20C3M (1961-2000);
-A2(SRES scenario) –
2046-2065
Purpose:
-to verify model ability to
simulate relative
frequency and number of
storm
- events (similarity of SLP
fields with corr>=0.85);
MPI-ECHAM5: Relative frequency of storm
events (left) and number of days (right)
Wind direction
frequency for
events when
daily V>=15 m/s
Monthly
frequency for
events when
daily V>=15
Wind speed over the
sea for events when
daily V>=15 m/s at
least in one grid
Спасибо за внимание!
А.В.Кислов
МГУ, географический факультет, кафедра метеорологии и климатологии
[email protected]