)d. 
ed 
ge 
els 
/as 
cht 
pal 
ric 
am 
od 
/as 
he 
ine 
cm 
ter 
M) 
for 
UQUORRS CH I XJ C) C XO S010 C2 C Z9 CH CI C2 C1 50 03. 5200 CI C9 12982 19 Q2 OLR 
A comparison of the WALDEM (figure 4) with the bathymetric 
DEM (figure 5) demonstrates the large similarity. The height 
differences between WALDEM and DEM on basis of pixel- 
bypixel comparison give a mean difference of -5 cm (WALDEM 
lower than DEM) and a standard deviation of 28 cm. 93% of the 
pixels are within the + 50 cm difference range, 73% within 
+ 30 em. 
The differences may attribute to: 
A. Water line procedure 
1) Georeferencing errors: not larger than 1 pixel (12.5 m) in 
horizontal 
2) Water line delineation: several pixels in transitional 
zone of the land and water due to a thin water film. 
3) Water surface modelling: smaller than 20 cm vertically. 
4) Selection of the interpolation method. 
B. Bathymetric DEM 
1) Echo sounding errors: approximately 20 cm in general 
and up to 40 cm over flats (Calkoen, et al, 1995). 
2) Omission of the topographic details due to interpolation 
between large spacing of bathymetric survey lines. 
4. CONCLUSIONS 
The water line procedure based on satellite radar imagery 
appears to be a reliable method for precision topographic 
mapping in the tidal range. 
The usefulness of SAR imagery for water line procedure is 
very much influenced by the wind and tide conditions at the 
time of acquisition. More than 6096 of the acquired ERS-1 
SAR images may be used for water line delineation. Wind 
velocities lower than 5 m/s often preclude a reliable 
discrimination between water and dry flat surface on the SAR 
image and the image can not be used. With a 35-day cycle (17- 
day sub-cycle), sufficient imagery will become available during 
the year for applying this method. 
The Global Classification Filter for segmentation of the 
speckled SAR image is proved effective by introducing the 
global concept. 
The water surface model (Wadden Model of RIKZ) for the 
acquisition time of the images is required and should be 
corrected on basis of actual gauging records and water surface 
curvature. 
The precision of the WALDEM compared with the bathymetric 
DEM are: for 93% of the pixels in the inter tidal area within 
+ 50 cm, and for 73% within + 30 cm vertically. 
The water line method is suitable for quantitative height 
measurements over the flats between high and low tide lines. 
By combination of the two methods —bathymetry over the 
channels and water line method over the flats— a considerable 
saving, more than 50% for this flat area, can be made on 
bathymetric surveying. 
The frequent SAR coverage allows for a monitoring on a more 
regular basis (once every or second year) for determination of 
the zones of erosion and sedimentation and the changes of the 
channels and gullies. 
767 
The present results are very encouraging and the study on the 
water line procedure may lead to an operational use of the 
method for topographic mapping in the inter tidal zone. 
ACKNOWLEDGEMENT 
The project was formulated under the Announcement of 
Opportunity for the utilisation of ERS-1 and J.ERS-1 data, 
issued by the European Space Agency (ESA) and the Japanese 
Space Agency (NASDA) respectively. 
The research project was carried out at ITC in cooperation with 
the North Netherlands Division, Ministry of transportation and 
public works (Rijkswaterstaat) who provided echo sounding 
data for bathymetric DEM and support in the water level 
measurement campaign. Thanks are due to the National 
Institute for Coastal and Marine Management (RIKZ), 
Rijkswaterstaat for providing the Wadden Model and the 
gauging records necessary for the water surface modelling. The 
assistance of A. Prakken, K.B. Robaczewska and R. van Dijk are 
kindly acknowledged. The filter is programmed by Mr. Li, Fei. 
Funding for the project has been provided by the Netherlands 
Remote Sensing Board (BCRS). 
REFERENCES 
Calkoen C. J., G. J. Wensink, J. Vogelzang and P. F. Heinen, 1995. 
EIBERS, Efficiency improvement of bathymetric surveys with 
ERS-1, BCRS report 95-01, Delft, 62 p. 
Koopmans B.N. and Y. Wang, 1995. "ERSWAD project- 
Measurement of land-sea transition from ERS-1 SAR at 
different phases of tidal water." Netherlands Remote Sensing 
Board, BCRS report 95-20, 64 p. 
Robaczewska K.B., RP. van Dijk, R. Plieger and M. 
Soerdjbali, 1991. "Een Model van de Nederlandse Waddenzee 
voor Berekening van Waterbewegingen Water Kwaliteit", 
GWAO-91.010,. 11 p,. 15 pl. 
Wang Y., B.N. Koopmans and D. de Hoop, 1995. ERSWAD 
project -Operation Logger, ITC report, (in press). 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996