Upload
garvey
View
19
Download
1
Tags:
Embed Size (px)
DESCRIPTION
Using fixed process-oriented model structure as a starting point for hydrological research Olga Semenova Hydrograph Model Research Group, St. Petersburg, RUSSIA, www.hydrograph-model.ru. 3. Red River of the North, MN, USA - South Branch of Buffalo River (749 km 2 ). 1. Background. - PowerPoint PPT Presentation
Citation preview
Using fixed process-oriented model structure as a starting point for hydrological researchUsing fixed process-oriented model structure as a starting point for hydrological researchOlga Semenova Olga Semenova
Hydrograph Model Research Group, St. Petersburg, RUSSIA, www.hydrograph-model.ruHydrograph Model Research Group, St. Petersburg, RUSSIA, www.hydrograph-model.ru
In this poster the general line of the Hydrograph model applications in different environments started at EGU 2009 is followed. The main idea is to show that a single process-based model with fixed distributed structure may be used to investigate the processes in very different environments. It seems reasonable in each case to adjust existing, widely tested model structure which is consistent with the scales of hydrological processes and the availability of parameter information.
As usual, some general thoughts on how we can advance hydrological modelling from its current state are presented as well:•The hydrological community should find the courage to leave some old, discredited theories and ideas in the past; instead of continuing to constructing model structures from non-physically-based blocks and conceptual simplifications; maintaining an understanding of the general processes combined with high-quality observational data may give a rise to new insights. •The convenience of some mathematical approaches should not take priority over the understanding of the processes. •New model structures should be carefully verified by independent data on similar watersheds (even if only runoff is available).•Model structure, where this is flexible, should not fully rely on calibration techniques to find “optimal structures”. Rational intellectual calibration of model parameters and decisions on appropriate model structures based on hydrological principles can be used if supported by an understanding of the underlying processes.•Reasonably developed, process-oriented, robust, fixed model structures can be an excellent starting point for hydrological research and will provoke serious consideration of the hydrological system and deficiencies in its description. For instance, the limitations of fixed model structures may point out new directions in observations that are necessary to more properly implement the model and form the basis of enhanced collaborations between modellers and experimentalists.
1. Background1. Background
2. Hydrograph Model2. Hydrograph Model
R • Single model structure for
watersheds of any scale
• Adequacy to natural
processes while looking
for the simplest solutions
• Minimum of manual
calibration
Forcing data: precipitation,
temperature, relative
humidity
Output results: runoff, soil
and snow state variables,
full water balance
Slope transformationof surface flow
Initial surfacelosses
Infiltration andsurface flow
Heat dynamicsin soil
Snow coverformation
Heat energy
Interception
Heat dynamicsin snow
Snow melt andwater yield
EvaporationWater dynamics in soil
Channel transformation
Runoff at basin outlet
Underground flow
Transformation of underground flow
PrecipitationRain Snow
3. Red River of the North, MN, USA - 3. Red River of the North, MN, USA - South Branch of Buffalo River (749 kmSouth Branch of Buffalo River (749 km22))
4. Dry Creek, Idaho, USA4. Dry Creek, Idaho, USA
5. California, USA – San Joaquin River 5. California, USA – San Joaquin River as an example (4341 kmas an example (4341 km22))
6. Summary6. Summary
AcknowledgementsAcknowledgements• Dry Creek project was made possible with funding from NOAA Grant NA08NWS4620047. The author Dry Creek project was made possible with funding from NOAA Grant NA08NWS4620047. The author
acknowledges the invaluable assistance from James McNamara and Pam Aishlin at Boise State acknowledges the invaluable assistance from James McNamara and Pam Aishlin at Boise State University, and Pedro Restrepo at the Office of Hydrologic Development of the National Weather University, and Pedro Restrepo at the Office of Hydrologic Development of the National Weather Service, USA.Service, USA.
• Red River study was made possible with contract from NOAA NWS OHD. The author acknowledges Red River study was made possible with contract from NOAA NWS OHD. The author acknowledges the assistance of Pedro Restrepo at the OHD NWS, the support of Mike DeWeese and Steve Buan the assistance of Pedro Restrepo at the OHD NWS, the support of Mike DeWeese and Steve Buan from the NCRFC, Carrie Ohlheiser from the NOHRSC, and Victor Koren from OHD. The administrative from the NCRFC, Carrie Ohlheiser from the NOHRSC, and Victor Koren from OHD. The administrative and management support of the contract by Robert Brown of Len Technologies is highly and management support of the contract by Robert Brown of Len Technologies is highly appreciated.appreciated.
• California project was made possible with funding from NOAA Grant NA09NWS4620044. The author California project was made possible with funding from NOAA Grant NA09NWS4620044. The author acknowledges the assistance from Xiogang Gao at the Regents of University of California, and Pedro acknowledges the assistance from Xiogang Gao at the Regents of University of California, and Pedro Restrepo at the Restrepo at the OHD NWSOHD NWS, USA., USA.
• Attendance to EGU 2011 was made possible with the support of German-Russian Laboratory for Attendance to EGU 2011 was made possible with the support of German-Russian Laboratory for Polar and Marine Research which is highly appreciated.Polar and Marine Research which is highly appreciated.
Catchment Area: 28 km2
Elevation Range: 1030-2130 m
Grasses, shrubs, and conifer forests vary with aspect and elevation
Low Elevation Grass
Mid Elevation Shrub
High Elevation Forest
Bare ground
Grass
Shrubs
Trees
Runoff formation complexes
01.200910.200807.200804.200801.200810.200707.200704.200701.2007
Te
mp
era
ture
, d
eg
ree
C
35
30
25
20
15
10
5
0
-5
vo
lum
e w
ate
r c
on
ten
t
0.25
0.20
0.15
0.10
0.05
0.00
observed simulated observed simulated
Lower Weather station (1151 m), Lower Weather station (1151 m), soil, 5 and cm depth, 2007-2008soil, 5 and cm depth, 2007-2008
01.200910.200807.200804.200801.200810.200707.200704.200701.2007
Te
mp
era
ture
, d
eg
ree
C
25
20
15
10
5
0
vo
lum
e w
ate
r c
on
ten
t
0.25
0.20
0.15
0.10
0.05
observed simulated observed simulated
observed simulated
05.200303.200301.200311.200209.200207.200205.200203.200201.200211.2001
sn
ow
de
pth
, m
1.0
0 .8
0 .6
0 .4
0 .2
0 .0
2002-2003
05.200902.200911.200808.200805.200802.2008
sn
ow
de
pth
, m
1 .4
1 .2
1 .0
0 .8
0 .6
0 .4
0 .2
0 .0
2008-2009
Tree Line station (1651 m), Tree Line station (1651 m), snow depth (m)snow depth (m)
Map showing location Map showing location of the watershedsof the watersheds
observedsimulated - even soi l depthsimulated - di str ibuted soi l depth
08.200906.200904.200902.200912.200810.2008
m3
/s
1. 2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Observed and simulated flow for even Observed and simulated flow for even and distributed soil depth casesand distributed soil depth cases
observedno losses, no cal ibrationno cal ibration, l osses implementedcal ibrated, losses implemented
08.200807.200806.200805.200804.200803.200802.2008
m3
/s
2. 1
1.8
1.5
1.2
0.9
0.6
0.3
0.0
Results of sequential calibration of Results of sequential calibration of groundwater parametersgroundwater parameters
observed simulated
01.200412.200311.200310.200309.200308.200307.200306.200305.200304.200303.200302.200301.2003
m3
/s
0. 9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
observed simulated
01.200512.200411.200410.200409.200408.200407.200406.200405.200404.200403.200402.200401.2004
m3
/s
1 . 7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Observed and simulated runoff, LowerGageObserved and simulated runoff, LowerGageNS (2000 – 2009) = 0.43 NS (2000 – 2009) = 0.43
observed simulated
12.1010.1008.1006.1004.1002.1012.0910.0908.09
So
il m
ois
ture
, V
C
0 .20
0 .15
0 .10
0 .05
0 .00
observed simulated
01.1111.1009.1007.1005.1003.1001.1011.0909.0907.09
Te
mp
era
ture
, d
eg
ree
C
2 4 .0
22 .0
20 .0
18 .0
16 .0
14 .0
12 .0
10 .0
8 .0
6 .0
4 .0
2 .0
0 .0
-2 .0
-4 .0
-6 .0
-8 .0
observed Hydrograph Sacramento
01.200910.200807.200804.200801.200810.200707.200704.200701.2007
m3
/s
70
60
50
40
30
20
10
0
observed Hydrograph Sacramento
12.200910.200908.200906.200904.2009
m3
/s
300
280
260
240
220
200
180
160
140
120
100
80
60
40
20
0
The hydrology of the study area is characterized by:
•snowcover modified by wind redistribution;
•high surface runoff from the major spring snowmelt event as a result of frozen soils;
•most rainfall occurring in spring and early summer from large frontal systems and the most intense rainfall in summer from convective storms over small areas;
•poorly-drained natural stream networks;•large amount of artificial agricultural
tiles influencing the redistribution of runoff into the streamflow network;
•small post-glacial depressions which may not drain to any natural external drainage system and have large storage capacity.
The study area is located at the south-east part of the Red River of the North watershed. The main landscape is agricultural fields. Annual precipitation amounts to 600 – 750 mm.
Observed and simulated soil moisture (50 cm depth) and Observed and simulated soil moisture (50 cm depth) and temperature at (10 cm depth) at Crescent Lake stationtemperature at (10 cm depth) at Crescent Lake station
Observed and simulated runoff by Observed and simulated runoff by Hydrograph and Sacramento modelsHydrograph and Sacramento models
Accounting for Depression StorageThe depth of depression storage is the part of surface runoff depth expended for filling micro-relief depressions of the watershed surface. Potential opportunities to retain a definite part of the surface runoff are caused by the free capacity of depression storage, i. e. by the difference between the maximum possible capacity (DM) and the amount of water in the depressions at the moment of commencing surface runoff (HP). Then, the depth of surface runoff retention (D) with known surface runoff depth (Hq) can be defined by the equation
M
qPM D
HHDD exp1)(
Accounting for artificial drainage systemThe winter soil frost creates conditions for surface flow when frozen soil infiltration rates are very low and prevent water from infiltrating deep into the soil. In those cases, surface flow occurs and goes directly to the agricultural drainage systems loosing some water to the depression storages on the way. In such a way, this type of flow can not be considered typical surface flow in the sense it is used in the Hydrograph model. This type of flow has a more deferred contribution into the river network. Given the available information about the river network and drainage system, which is more generic rather than specific, accounting for those processes was made possible by a change in the surface runoff elements parameters. In this way, the values of the conditional constants a* (see [1] for details), were changed from 1000 to 10 for a surface type of flow and from 100 to 5 – for soil type of flow.
Used precipitation dataUsed precipitation data
1) Station data2) Snow Data Assimilation
System (SNODAS)3) MAPX – mean areal
precipitation calculated using the radar- and gauge-based information.
California
Feather River
American River
San Joaquin River
Ü Ü4237
119
Ü15 - 23
23 - 33
33 - 43
43 - 51
51 - 63
ÜEvergreen forest land
Lakes
Mixed forest land
Residential
Shrub and brush tundra
Ü
s1061
s1062
s1063
s1065
s1066
s1067
s1068
s1070
s1073
s1074
s1075
s1117
s1118
s740
s746
s747
s748
s751
s752
s818
s8369
s845
#0
#0
#0
#0
#0
#0
#0
#0
#0#0
#0
#0
#0
#0#0
#0
#0
#0
#0
#0
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
%2
FLR
MBR
TMR
PSR
PCK
OKH NTP
NFR
MHP
KSP
HNT
GTM
GRV
GRM
FRT
CRR
CHM
BGC
BDM
AGW
MTF
BTT
NFR
TMR
PSR
KSP
GTM
GRV
GRM
FRT
VLC
UBC
STR
RCK
GEM
FCH
AGW
Ü
San Joaquin River watershed
#0 precipitation
!( temperature
%2 relative humidity
San Joaquin River San Joaquin River watershed elevation watershed elevation
schemeschemeSan Joaquin River San Joaquin River
precipitation distributionprecipitation distribution
San Joaquin River soil San Joaquin River soil type distributiontype distribution
San Joaquin River land San Joaquin River land use type distributionuse type distribution
San Joaquin River San Joaquin River meteorological stations meteorological stations
distributiondistribution
simulated observed
10.200508.200506.200504.200502.200512.2004
m3
/s
850
800
750
700
650
600
550
500
450
400
350
300
250
200
150
100
50
0
simulated observed
10.200608.200606.200604.200602.200612.2005
m3
/s
850
800
750
700
650
600
550
500
450
400
350
300
250
200
150
100
50
0
simulated observed
12.200710.200708.200706.200704.200702.200712.2006
m3
/s
200
150
100
50
0
simulated observed
10.200808.200806.200804.200802.200812.2007
m3
/s
300
250
200
150
100
50
0
Observed and simulated runoff for San Joaquin River (2005 – 2008)Observed and simulated runoff for San Joaquin River (2005 – 2008)No calibration is implementedNo calibration is implemented
•The Hydrograph model performs well in different environments•Most parameters are estimated from physical characteristics and do
not require further calibration•Some parameters of conceptual components do require calibration•The correct way to approve or reject a model structure is by testing it
multiple times over basins in different conditions•Reasonably developed, process-oriented, robust, fixed model structure
can be an excellent starting point for any hydrological research provoking
serious consideration of the hydrological system and deficiencies in its
description
We do not have resources for developing a specific model
for every single watershed… We do not have resources to
play “modelling exercises” in the sense of making
thousands of runs aimlessly searching for some hidden
model structure or parameter meaning….
Can YOU afford them?