ome 
ugh 
ly to 
the 
are 
ghly 
off, 
odel 
1 be 
the 
) or 
rage 
age 
fing 
aust 
| of 
per 
eful 
za (DIC) 
tes 
tal 
Bertens, Jurjen 
  
form. As more detail is needed for the analysis, these maps are being revised. Detailed precipitation records are available for 
both watersheds (figure 4). 
The relationships between the input described above are expressed as mathematical expressions. Input from precipitation 
will reach the surface. Water can the infiltrate or remain on the surface. Infiltration is calulated according to the Green & 
Ampt model (equation 1, Amaru’ Michele, 1995). 
Q:F-K:AD, JI FK AD? -8-K-ArQr-A0 € F)] (equation 1) 
2 2 
Potential infiltration is calculated for each timestep as a function of hydraulic conductivity (K), cumulative infiltration in 
previous timesteps (F), the suction head at the wetting front (y) and change in moisture content (/0). Potential infiltration is 
used as value for infiltration if the storage capacity of the soil allows infiltration. Water that can not infiltrate (either due to 
limited potential infiltration or limited storage capacity), will remain on the surface, after which it will move through the 
watershed according to either channel flow or overland flow. 
  
AF = 
  
Channel flow is modelled as a kinematic wave (without accelerations due to the characteristics of flow itself). Flow velocity 
can the be calculated with the Darcy-Weisbach equation (equation 2) 
v = La R-S (in which Fs — Chezy Nail) (equation 2) 
where g is gravity acceleration, f is the Darcy-Weisbach friction factor, R is the hydraulic radius and S is the slope. When 
Chezy's C is regarded equal to R'°/n, Manning's equation for the calculation of open channel flow velocity is obtained 
(equation 3), 
243 142 
RS 
n 
where is the Manning's roughness coefficient, which depends on land cover. 
y (equation 3) 
Overland flow will occur on a sloping surface. Three possible situations can be identified. First, a surface with a hydraulic 
conductivity smaller than the precipitation intensity, (infiltration excess), secondly saturation from above (saturation), and 
thirdly saturation from under (exfiltration). Equation 4 is the kinematic wave equation for overland flow, 
BX, 5, 1)5 0Y (5, 1) (equation 4) 
woo rDI Gu) = = = 
Where W., is effective precipitation (here precipitation minus infiltration), m is a factor (set at 0.5), U is discharge form a 
pixel and Y (s,t) is the depth of flow. Solutions for equation 4 can be found using the method of characteristics. 
  
Values for these parameters are updated after each time interval. Output for each time interval they are used as inputs for the 
aaplication of the soil erosion model, the output of which consist of estimates of sediment production. This output, in turn, is 
used for the calculation of suspended sediment load. The latter parameter can be measured during and after construction for 
model calibration and validation. 
6 CONCLUSIONS 
The approach presented provides a means to make certain predictions concerning the likely modeifications of hydrology 
related processes as a consequence of changes due to infrastructure construction and operation. 
The method described, based on the sequential application of a series of dynamic, distributed models, will produce a final 
output in terms of measurable parameters, significant for EIA (channel discharge, soil loss, sediment load). These 
parameters can be determined to test and calibrate the model both prior to and during construction. 
Minimum inputs required for the model are precipitation records, digital elevation, a soil and land cover map. These inputs 
are often available or can be obtained relatively easily. The only insurmountable difficulty is, obviously, lack of precipitation 
data. 
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000. 181