Bertens, Jurjen 
  
important in some small streams and local increases in flood hazard may occur. Increased water-logging problems in some for] 
areas may also appear as a consequence of landform changes. These changes may affect erosion and related effects, although bot! 
probably not as much as during the construction phase. On the other hand, pollution of surface and groundwater is likely to 
be more important at this stage. Combustion pollutants and leaks from vehicles will affect soil and groundwater in the The 
vicinity of the motorway and surface waters downstream from it. In figure 3 the impacts resulting from construction are wil 
considered. Am 
As shown in figure 2, soil erosion has been identified as a potentially significant impact (3). The Deba river is highly 
polluted and additional pollution is not a major concern. Relevant processes (4) are all processes related to erosion; runoff, 
   
  
   
  
infiltration, surface roughness etcetera. Pot 
pre 
Fundamental relations between these processes are conceptually established (5) and transformed into a mathematical model use 
(6), where relations are represented by equations and parameters represent field characteristics. If all model input can be lim 
obtained the model is calibrated and validated (8). Thus, relevant processes are formulated in a way that complies to the wat 
limits of data availability (Maidment, 1996). e 
If not, conceptual and/or mathematical changes (simplification of interrelations) will have to be introduced (775,6) or can 
reasonable estimates can be used. In this case study e.g. no data on storage capacity of soils is available. Thus, storage 
capacity can be estimated as a function of soil type. In (9) surface characteristics for the location of the planned project are 
altered for post-project conditions. Since no detailed information on slope modification can be obtained at pre-project stage 
(and will probably occur at sub-grid scale), adjustments in the model will have to be made. Since no means of calibrating 
and validating the output of the model with respect to the effects of infrastructure exist, alternative validation methods must wh 
be used (10). A possible way of calibrating the model, especially during the construction phase, is the determination of Ch 
suspended load in streams. As a last step the model’s predictions are translated from soil loss values (gram per unit area per (eq 
time interval) to relevant terms which allow for interpretation and comparison (e.g. “equivalent” agricultural land loss, useful 
life of a reservoir due to siltation). 
wh 
5.3 Model application 
Ov 
scati ; cot 
ve Aon Sr e NN Average annual precipitation Basque Country is thi 
: : (1998-1999) A £e 
elevation model (as detailed as 
possible), rainfall data (time series at . 
hourly timescale), soil characteristics 
and land cover. The final model output Wi 
will be, as mentioned above, stream pix 
discharge and suspended sediment 
load (expressed in terms that are Va 
significant for EIA). aar 
Discharge records with sufficient Use 
detail are available two small Inc 
watersheds within the Deba watershed 
(Urkulu; 7.5km° and Aixola; 4.7km°). ccocazafw +) AE à. 
These have been selected as 00 PR Jatt ut adil 6 ( 
: : d d A z^ = , > / 9 Ramgauge Diputación Foral de Gipazhoa (DF 
particularly suitable for the initial y D Fsinpuge Iretétaio Hacineal Metemalor! 
  
   
application, calibration and validation ; 3X Raingange DFG-Gchiermo Vasco Th 
of the model (figure 4). Hun ad [] Watershed of the Deba River rel 
+. : S Tet Th 
A digital elevation model is available Figure 4. Annual precipitation (mm/year) and raingauge stations for the Basque oul 
for the entire study area with a coarse Country. Snai 
: try, Spain, ai 
resolution (25 meter contour p 
intervals). Since this is not Mi 
considered to be sufficiently accurate, a more detailed digital elevation model is being prepared for the mentioned test sites ar 
(with contour intervals of 5 meter). Soil and land cover maps, originally made at the 1:25.000 scale, are available in digital da 
  
180 International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000.