Experience with modelling of runoff formation processes at basins of different scales using data of representative and experimental watersheds

1
Experience with modelling of runoff formation
processes at basins of different scales using
data of representative and experimental
watersheds

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

2
Introduction

• Models need parameters values
• The problem of heterogeneity
• Example infiltration coefficient of the upper
  soil layer
• Idealized representative slope
• The problem of calibration

Areal extent, m2 Cv
10-3 (filtration tube) 10
10-1 (field filtration device) 1
102 (sprinkling-machine) 0.1
105 (elementary watershed, estimation in inverse way by observations of precipitation and surface runoff) 0
3
Objectives
to demonstrate that using observational data of
small jointly with the appropriate modelling
algorithms gives the possibility to avoid the
calibration procedure and transfer estimated
parameters (without change for a given landscape
zone) to other basins, including those with
scarce availability of information.
4
Main principles of model development

• Universality (response to PUB challenge)
• Balance between simple solutions and adequate
  description of natural processes
• Apriori estimation and systematization of main
  parameters (without calibration for any new
  object)
• Routine forcing data

5
Deterministic Modelling Hydrological System
(DMHS or model Hydrograph, by Prof. Yu.B.
Vinogradov)
6
DMHS features

• Distributed
• Calculating interval 24-hour or less
• Forcing data precipitation, temperature and
  humidity
• Output runoff hydrograph, water balance
  elements, state variables of soil and snow cover

DMHS key concepts

• Concept of runoff formation complexes
• Concept of runoff elements (see for details
  Vinogradov 2003, 2008)

DMHS parameters

• Soil properties
• Vegetation cover properties
• Slope surface
• Underground water
• Climate parameters

7
The spatial-computational schematization of the
basin
8
What do we need from small watersheds?

• Observational data on representative basins to
  calibrate some model parameters
• Evaluation and systematization of the
  representative landscape properties (i.e. apriori
  assessment of model parameters)

What do we need from experimentalists?

• Understanding of the processes and its clear and
  proved explanation
• Understanding of the models and their objective
  and active evaluation

Mutual interaction between modellers and
experimentalists
9
Study objects
Valday experimental station (research is still in
progress)

Suntar-Hayata range geophysical station



Nizhnedevitskaya Water Balance Station
Kolyma Water Balance Station

Mogot experimental plot
Mild CLIMATE Extreme
Plain, hilly RELIEF Mountainous
Steppe LANDSCAPE Tundra, taiga
Seasonal PERMAFROST Continuous
10
PRELIMINARY RESULTS
I. Nizhnedevitskaya water balance station
SOIL TEMPERATURE
0.2 m
0.8 m
11
SNOW CHARACTERISTICS
Snow height at Nizhnedevitskaya observational
station
12
SOIL MOISTURE
Stream Dolgy, area 2.51 km2, content of moisture
in 1-m layer
13
RUNOFF
Devica river at Tovarnya, area 103 km2
Sosna river at Elec, area 16300 km2
14
II. Kolyma water balance station
SOIL TEMPERATURE
0.4 m
0.8 m
15
RUNOFF
Yuzhny stream, area 0.27 km2
16
Detrin at Vakhanka river mouth, area 5630 km2
17
Kolyma at Kolymskoye, basin area 526000 km2
18
Suntar at Sakharynia river mouth, area 7680 km2
III. Suntar-Hayata range experimental station
19
Yana at Dgangky, area 216000 km2
20
Nelka at Mogot, area 30.8 km2
III. Mogot experimental plot
21
Katyryk at Toko, basin area 40.2 km2
22
Timpton at Nagorny, area 613 km2
23
Uchur at Chyulbu,area 108000 km2
24
Statistics on observed vs simulated flow
(averaged for all basins in Eastern Siberia)
Daily Year
Nash-Sutcliffe 0.78 0.93
Relative error (in absolute value) 36 10
25
Conclusions

• The results aim to demonstrate the possibility of
  a single hydrological model application for
• runoff simulations at large-scale basins, as well
  as for fine time step representation of
  individual hydrological process at the local
  scale
• simulation at various landscape and climate zones
  with different driving processes.

Desired future
The observations should be carried in tight
interaction with the development of hydrological
models, i.e. the experiments and observations
schemes and components are to be coordinated in
order to be used in the adoption or rejection of
current hydrological theories and assumptions.
26
REFERENCES (in English)

• Vinogradov, Yu.B., 2003a, River Runoff Modeling
  in Hydrological Cycle, edited by I.A.
  Shiklomanov, in Encyclopedia of Life Support
  Systems (EOLSS), Developed under the auspices of
  the UNESCO, Eolss Publishers, Oxford, UK,  
  http//www.eolss.net
• Vinogradov, Yu.B., 2003b, Runoff Generation and
  Storage in Watershed in Hydrological Cycle,
  edited by I.A. Shiklomanov, in Encyclopedia of
  Life Support Systems (EOLSS), Developed under the
  auspices of the UNESCO, Eolss Publishers, Oxford,
  UK,   http//www.eolss.net