Deterministic-stochastic modelling as a potential tool for the assessment of climate change impacts on hydrological regime in polar regions

1
Deterministic-stochastic modelling as a potential
tool for the assessment of climate change impacts
on hydrological regime in polar regions

• Olga Semenova
• State Hydrological Institute
• St. Petersburg, Russia

2
Objectives
The goal of the research is to develop a tool for
assessment of the possible changes in the
annual, seasonal and extreme runoff
characteristics of watersheds of Eastern Siberia
on the base of deterministic-stochastic modelling

• Necessary conditions
• Process-oriented deterministic model with
  physically observable parameters, minimum of
  calibration and ability to port calibrated
  parameters to other watersheds
• Stochastic model providing the meteorological
  inputs according to climate change projections

State Hydrological Institute, St. Petersburg,
Russia
3
Research strategy
Physically observable parameters
Parameters of observed daily meteorological series
Simulated scenarios of daily meteorological data
Deterministic hydrological model
Stochastic Model of Weather
Runoff generation processes simulations
Climate change projections
Numerical evaluation of runoff characteristics
changes in probabilistic mode
State Hydrological Institute, St. Petersburg,
Russia
4
Study area
State Hydrological Institute, St. Petersburg,
Russia
5
Hydrograph model (deterministic modelling)

• Single model structure for watersheds of any
  scale
• Adequacy to natural processes while looking for
  the simplest solutions
• Minimum of manual calibration

R
Forcing data precipitation, temperature,
relative humidity Output results runoff, soil
and snow state variables, full water balance
State Hydrological Institute, St. Petersburg,
Russia
6
Weather model (stochastic modelling)

• Simulation of daily precipitation, temperature
  and relative humidity
• Simulation of annual and intra-seasonal
  variations
• Simulation for hexagonal system of representative
  points
• Temporal correlation of meteorological elements
• Spatial correlation of meteorological elements
• Parameters are estimated from observed series of
  meteorological data
• Parameters may be modified according to climate
  change projections

R
State Hydrological Institute, St. Petersburg,
Russia
7
DM state variables
0.8 m depth
0.4 m depth
Soil temperature at different depths
Kolyma Water balance station
Snow characteristics
State Hydrological Institute, St. Petersburg,
Russia
8
DM small and middle scale basins
---------- simulated ---------- observed
Timpton at Nagorny, basin area 613 km2
Vitim at Bodaybo, basin area 186000 km2
State Hydrological Institute, St. Petersburg,
Russia
9
DM Large-scale basin
Lena at Kusur basin area 2,4 million km2
---------- simulated ---------- observed
State Hydrological Institute, St. Petersburg,
Russia
10
SM annual values
Chara meteorological station. Annual
precipitation (mm)
Calc Obs
mm
Exceedance probability,
State Hydrological Institute, St. Petersburg,
Russia
11
SM correlation of annual values
Spatial correlation of annual temperature, Lena
river basin
State Hydrological Institute, St. Petersburg,
Russia
12
SM seasonal values
Monthly distribution of precipitation
Monthly distribution of temperature
Bodaybo station
Vostochnaya station
State Hydrological Institute, St. Petersburg,
Russia
13
SM daily values
Daily precipitation, Suntar-Hayata station
State Hydrological Institute, St. Petersburg,
Russia
14
SM correlation of daily values
Spatial correlation of daily temperature, Lena
river basin
State Hydrological Institute, St. Petersburg,
Russia
15
DSM maximum flows
Timpton river at Nagorny, 613 km2
State Hydrological Institute, St. Petersburg,
Russia
16
DSM daily flows
Ebetiem river at Ebetiem, 114 km2
State Hydrological Institute, St. Petersburg,
Russia
17
Conclusions

• The deterministic hydrological model Hydrograph
  performs well in arctic region at different
  landscapes and scales
• Most parameters are estimated from physical
  characteristics and do not require any
  calibration
• The stochastic model takes into account annual,
  seasonal and daily variation of meteorological
  elements and their spatial and temporal
  correlation

Next step would be
generation of ensembles of precipitation and
temperature forcings according to IPCC climate
change projections to obtain probabilistic
estimates of annual, seasonal and daily extreme
hydrological variables for large scale basins
State Hydrological Institute, St. Petersburg,
Russia
18
Acknowledgements

• The research is being conducted within the
  research grant funded by the German-Russian
  Otto-Schmidt laboratory for Polar and Marine
  Research
• The IPY Oslo Stipend granted by the Research
  Council of Norway is appreciated

State Hydrological Institute, St. Petersburg,
Russia