696 
500m 
moraine material 
dolomite rubble uith^ 
shallow episodic 
weathered marly flow system 
tuffs with shallow 
permanent flow 
system spring 
area 
dolomite rock 
with deep per 
manent flow 
system 
complex of 
flou systems 
in; 
hydrological land 
units 
flow systems 
(legend; Fi- 
qure 5) 
marly tuffs 
high grasslands on 
HOF, SOF, SSF 
weathered marly tuffs 
open pine forest on 
weathered marly tuffs 
bare marly tuffs 
wet grass species in 
seepage zone 
and GF(SP) 
dolomite 
shrubs on rubble 
HOF, SOF, G F ( SE 
rack/rubble 
bare rubble 
dense pine forest 
on moraine material 
open pine forest 
on thin soils 
bare dolomite rock 
spring area 
and GF(DP) 
Figure 6. Hydrological land units and 
flou systems of marly tuffs and dolomite 
rock/rubble. 
the deeper bedrock, is either covered with shrubs, 
or completely bare. Generally, the water discharges 
as structurally controlled concentrated springs. 
Horton overland flow and saturation overland flow 
are the storm runoff mechanisms. 
High grasslands with some open pine forests domi 
nate the vegetation of the marl unit. Due to the 
mass-movements in the residual overburden the un 
weathered bare rock is locally exposed. Here a 
Horton type of overland flow may occur. The small 
amount of groundwater flow can be explained by a 
shallow permanent system in the weathered marls. In 
the seepage zones of this unit many wet grass 
species are found. Saturation overland flow and sub 
surface storm flow in the soil layer are important 
runoff processes. 
The first effort towards a digital classification 
of the main vegetation and landuse units was per 
formed in March/April 1986. Therefore a Landsat TM 
tape of August 1984 and two Landsat MSS tapes of 
August and October 1984 have been analysed. Several 
supervised classification methods have been tested : 
1. Two band box classifier (MSS 7-5 and TM 4-3, 
4-5 and 5-3). 
2. Maximum likelihood classifier using all TM 
bands. 
3. To eliminate the shadowing effect a fast paral 
lelepiped two-features classification has been exe 
cuted, using : 
*combinations of TM ratios 4/5, 3/4 and 3/5. 
*"greenness" and "yellowness factor" (leaving out 
the "brightness factor") of a principal component 
transform of Landsat TM bands. 
4. Multitemporal classification using the near 
infrared MSS bands of August and October 1984. 
A final evaluation of the classification results is 
scheduled for the end of 1986. The results will be 
compared to a map, aerial photograph (1:25.000) and 
field data-based classification. 
Hydrograph analysis of the 1985 and 1986 dis 
charges of the subcatchment of the Upper-Boite area 
will provide the parameters of the groundwater flow 
systems. During the spring and summer of 1986 the 
field plot measurements of interception and soil 
moisture storage capacity, maximum percolation rate, 
etc. will be sampled for each hydrological land unit. 
Although it is premature to give definite reflect 
ance indices which may be correlated to hydrological 
parameters, at this stage combinations of ratios of 
MSS 5/7 and TM 3/4, 4/5, 3/5, "greenness", "yellow 
ness" and "brightness factors" and the thermal infra 
red values seem to be promising as hydrologically 
important features. It should be realized that only 
for physically-based models in which the hydrological 
parameters are uniquely related to landcover charac 
teristics, such a correlation with Landsat reflect 
ance indices is possible. 
3.3 Verification of the flow model in several control 
areas 
The remote sensing-supported hydrological mapping and 
modelling will be tested in several control catch 
ments. Two catchments are situated to the west and 
one to the east of the Upper-Boite catchment (Figure 
4) . 
4. CONCLUSIONS 
As the existing catchment models are of limited ap 
plicability in heterogeneous (Alpine) environments 
and are unsuited to incorporate remotely-sensed data, 
a new generation of models should be developed. A 
possible outline of such a new modelling approach is 
described in this paper. The methodology is focused 
on two items : 
1. The identification of the spatial distribution 
of hydrological units and processes is strongly sup 
ported by satellite imagery. 
2. The basic input parameters are physically-based 
which makes it possible to relate field data-based 
parameters to Landsat reflectance indices. 
In the near future this remote sensing-supported 
semi-distributed model will be applied to several 
catchments in the N-Italian Dolomites. 
Hendriks, M 
ogical da 
Free Univ 
Holtan, H.N 
N.C. Lope 
watershed 
United St 
Washingto 
Meyerink, A 
survey of 
ITC/GUA/V 
Mosley, M.P 
hydrologi 
Peck, E.L., 
Strategie 
hydrologi 
Flight Ce 
Peck, E.L., 
Suitabili 
use in hy- 
Hydromete 
Ragan, R.M. 
using Lan^ 
Hydraulic 
15387, 10< 
Rango, A. 1' 
to hydro li 
Vol. 21, i 
Seyhan, E., 
1985. Mul- 
of the hy< 
Northern : 
no. 7 : 1( 
Simmers, I. 
approach 1 
J. Hydrol. 
Soil Conser'' 
ing Handbc 
ment of Ac 
REFERENCES 
Crawford, N.H. and R.K. Linsley. 1966. Digital simu 
lation in Hydrology : Stanford Watershed Model IV. 
Technical Report no. 39, Department of Civil 
Engineering, Stanford Univ., Standford, CA. 
Ebisemiju, F.S. 1979. An objective criterion for 
selection of representative basins. Water Resources 
Research, 15(1) : 148-158. 
Engelen, G.B. 1963. Gravity tectonics in the N.W. 
Dolomites. Geologica Ultrajectuna No. 13, 
Rijksuniversi.teit Utrecht. 
Engelen, G.B. 1974. Hydrogeology of the Sasso Lungo 
Group. A Dolomitic Reef Stock in the Alpine Dolo 
mites of North Italy. J. Hydrol. 21 : 111-130. 
Engelen, G.B. 1984. Hydrological systems analysis. A 
regional case study. Report OS 84-20. Institute of 
Applied Geoscience TNO-DGV, Delft. 
Engman, E.T. 1982. Remote sensing application in 
watershed modeling. In Applied modelin in catch 
ment hydrology, Proc. of the Int. Symp. on rain 
fall-runoff modeling. Water Resources Publ. : 
473-494. 
Fliri, F. 1975. Das Klima der Alpen in Raume von 
Tirol. Monographien zur Landeskunde Tirols. 
Universitätsverlag Wagner, Innsbruck-München. 
Fischer, G.T. and J.P. Ormsby. 1982. The application 
of remotely sensed observations to hydrologic 
models. In D.N. Body (ed.), Application of results 
from representative and experimental basins : 
409-428. 
Groves, J.R. and R.M. Ragan. 1983. Development of a 
remote sensing based continuous streamflow model. 
In Proc. of the 17th International Symposium on 
remote sensing of environment : 447-456. Ann Arbor, 
Michigan .•