SUBSEQUENT PROCESSING OF DEM RESULTS 
DEMs for further processing were available with the 
parameters listed in table 1. A five point margin was 
cut from the models to eliminate marginal effects that 
might be caused by the surrounding frame, thus result- 
ing in 45,600 elevation points on 0.182 m? in the labor- 
atory models (and 102,400 from 0.92 m° in the field 
models). For each microrelief two characteristic values 
were computed, i.e. the indices RRC (random roughness 
coefficient, according to Currence & LoveLy 1970) and 
the ratio TSA/MA (total surface area to map area, this 
helps providing a quantitative measure for the raindrop 
density). 
  
  
Parameter Laboratory Field 
Area of measurement 0.2 m? 0.98 m? 
Spacing 2mm 3mm 
Height resolution 0.2 mm 0.2 mm 
Number of data 50,000 108,900 
  
  
  
Table 1: Parameters for microrelief measurement 
RRC represents the variance of the smoothed height 
values. Smoothing results in elimination of external 
relief effects like slope and directional relief elements. 
TSA/MA was computed as ratio of the sum of all surface 
areas of the DEM (there are 45,171 resp. 101,761 raster 
squares) divided by the DEM ground area (i.e. the TSA 
mapped on the ground). For the registration of spatial 
structures in the microreliefs semivariograms were com- 
puted. These data were smoothed too, in order to get 
out trends and the low-frequency variance caused by 
slope. 
DETERMINISTIC MODELS 
Effective input of rainfall energy 
For calculation of the effective rainfall energy that hits 
the soil surface, the total surface area (explained above) 
and as well the normal component of rainfall energy, 
depending on the angle of drop impact, is needed. 
This parameter is computed by determining the slope 
of the soil surface in each grid square of 4 mm? for lab 
and 9 mm! for field plots. Assuming vertical rainfall 
and a horizontal soil surface, the normal component 
of the impact energy corresponds to the total kinetic 
energy. With increasing slope at the striking point the 
normal component decreases with respect to the impact 
vector, leading to a decrease in effective energy. The 
normal component thus is proportional to the cosine 
of the slope at the impact point (fig. 2). 
Effective input of kinetic rain energy with respect to 
TSA and amount of rainfall may be expressed as: 
E.eff = ma/tsa * X (cos a * E.kin) (1) 
a = soil surface slope; ma = map area (m°); tsa = total surface area (m?) 
    
   
   
  
  
  
  
  
  
  
      
    
    
     
    
    
   
  
  
  
  
  
  
  
  
  
   
  
    
    
   
   
   
   
    
   
  
  
  
  
  
  
  
  
   
    
    
  
    
e.kin 
     
  
    
surface area 
map area 
  
Fig. 2: Relation between surface slope, impact angle 
and the distribution of impact force 
o = slope angle, E.kin = kinetic energy of raindrop, n = normal component 
of impact force, t = tangential component of impact force 
Depressional storage 
Storage capacity was determined by finding each depres- 
sion in the relief and estimating its circumference, sur- 
face area and volume. Identification of a depression 
was carried out starting from the pour point, defined as 
the lowest point between two or more depressions and 
within an elevated area. With the pour point as basis, 
closed contour lines were determined that laid either 
at the same or at a lower height than that of the pour 
point. The area within such a contour line was defined 
as a depression. Thus the extent of a depression is 
equal to the length of the respective contour line. The 
vertical distance between the lowest point and the pour 
point is the maximum filling height. The surface area 
was calculated using vector geometrics the same way 
as for the TSA. Elevated areas within a depression (is- 
lands) were subtracted. This technique of finding depres- 
sions complies with the theory proposed by Uuau & 
Dickinson (1979) and Huanc & Braprorp (19908). How- 
ever, they began from a local depression and identified 
the pour point by checking the list of neighboring 
points until finding a point with a lower elevation. 
This leads to more or less rectangular depression forms. 
The procedure proposed here has the advantage, that 
the whole depression is characterized independently 
of its form. 
RESULTS AND DISCUSSION 
Microrelief characterization 
The microrelief parameters obtained from the DEMs 
are summarized in table 2. 
In all cases, rainfall lowers the TSA/MA ratio by about 
0.1. Thus, the total surface area of the three microrelief 
treatments differs significantly and the relative change 
caused by rainfall is slightly greater for the fine micro- 
relief. In the field experiments the ratio of TSA/MA for 
both sites showed values that were comparable to that 
from the lab. Thus, the laboratory simulations seem to 
be representative of natural conditions as measured in