PUB in practice: case studies - Runoff Prediction in Ungauged Basins: Synthesis across Processes, Places and Scales - page 351

Table 11.20. Module parameters

Module Parameter Symbol Extent Range

BTOP (runoff generation) Discharge decay factor m Block 0.01 ~ 0.1

Groundwater discharge ability coefficients D sand , D silt , D clay Basin 0.01 ~ 2.0

Soil freezing threshold temperature ( C)

T t Basin 0.0 ~ 2.0

Root zone soil moisture capacity (mm) S rmax Grid cell 50 - 1500

Block average Manning ' s coefficient n 0 Block 0.01 ~ 0.8

Snow accumulation threshold temperature ( C)

- 2.0 ~ 2.0

Snow Basin

Degree-day index Basin 1.0 ~ 1.9

Snowmelt threshold temperature ( C)

Basin 0.1 ~ 1.0

Re-freezing coefficient Basin 0.01 ~ 0.09

and calculated for a grid cell using an explicit relationship
(Lu et al., 1989 ). The three-dimensional physiographic
heterogeneity of the basin was considered mutually in terms
of topography, soil types, vegetation cover and root depth.
In this application we only use the runoff generation, snow
and potential evapotranspiration modules, and the related
module parameters are shown in Table 11.20 .
Discharge decay factor (m) was assumed to be homoge-
neously distributed inside a block. Assuming that channel
slope is proportional to the resistance of flow, BTOP uses
the following equation to calculate the Manning

soil map). The soil freezing threshold temperature of a block
was also taken from the literature based on the major soil type
present in the block. Typical values from the literature were
also used for the snow module parameters (set for the whole
basin) and hence slight tuning may be required. No evapo-
transpiration model parameters were tuned as they were all
taken from the literature (Zhou et al., 2006 ).

Results
The overall validation (1977 2000) performance of YHyM

at the outlet of each block is good (Nash

-

' s rough-

- Sutcliffe scores

ness coefficient (n) of a grid cell:

of 0.7 0.84). The simulated and observed hydrographs at

several locations within the MRB are shown in Figure
11.84 , and it can be seen that low flows are simulated quite
well. However, peak flows are sometimes overestimated,
and it is thought that this is due to an insufficient number
of rain gauge stations (only 65 for the whole MDB, and
only one for Block 8 and none for Block 7) which reduces
the ability to realistically replicate the spatial variability of
rainfall over the MRB.
To demonstrate the applicability of YHyM to PUB,
Figure 11.85 shows daily average catchment character-
istics at all points in the MRB for an exemplar wet and
dry month. Rainfall in areas with no rain gauge stations
was approximated using data from nearby stations and
applying the Thiessen polygon method. At times (e.g.,
high precipitation), this approximation was unrealistic
and hindered the model performance to some extent.
However, the model

-

1 = 3

tan β

n ¼ n 0

tan β 0

where β is the local channel slope and β 0 is the block average
slope. The groundwater discharge ability (D 0 )ofagridcell
was calculated following Hapuarachchi et al.( 2004 )as

D 0 ¼ U clay D clay þ U sand D sand þ U silt D silt

where U clay , U sand and U silt are the percentages of clay,
sand and silt present in a particular grid cell, respectively.
It is assumed that soil texture inside a grid cell is homoge-
neous; thus D clay , D sand and D silt could represent additional
soil textural properties (particle size, pore size, etc.). These
parameters are defined for the whole basin and slight
tuning is needed. The maximum storage capacity of the
root zone (S rmax ) for each grid cell was calculated as

s performance at simulating some
of the internal points with limited data (i.e., Thoeng,
Ubon and Yasothon) was found to be satisfactory, sug-
gesting YHyM is a useful tool in ungauged (or sparsely
gauged) basins. Further research is required to investi-
gate the details as to why, where and when YHyM
performs well or poorly in the MRB (e.g., elevation
effects, density of gauges, uncertainties associated with
models inputs, uncertainty associated with observed
runoff etc.).

'

S rmax ¼ RD ðθ fc −θ wp Þ

where RD is the depth of the root zone of a grid cell (obtained
from literature based on the land use map);

θ fc is the field

capacity (m/m); and θ wp is the topsoil moisture content at

wilting point (m/m). Literature values (Rawls et al., 1982 )of
θ fc and θ wp for each grid cell were defined based on the soil

textural properties (from Food and Agriculture Organization