GIS AND REMOTE SENSING TECHNIQUES FOR THE INVESTIGATION OF COASTAL DYNAMICS: THE GRADO ISLAND CASE

Andrea Favretto (1), Giovanni Mauro (2) - Luciano Lago (1)

(1) Dept. of Geographical and Historical Sciences - University of Trieste
(2) CETA (Gorizia)
The present study has been realized with the full collaboration between the Authors. However, for open competitions, the analysis of the Franceschino Land Register map is due to Luciano Lago, the elaboration of satellite data is due to Giovanni Mauro, the rest of the paper to Andrea Favretto.

Introduction
In this paper we have tried to outline the evolution of the coastline around Grado town in the northern Adriatic sea. The studied area is: the coastline from Grado island to the Isonzo river mouth and the close inner area of Grado municipality.
Grado is one of the most important summer centres in the Friuli - Venezia Giulia region and also in the Northern Adriatic Sea. It is well known in Italy and abroad for several reasons: its large and equipped beach, its commercial and cultural attractions. Grado island and its lagoon are very significant both for economic (tourist activities), and also environmental reasons (Ciceri, 1980).
The coast morphology in the studied area is in part due to the actual river terrain contributions and in part to the old morphology of this territory. This last was delineated by the delta mouth of the ancient Natisone, Isonzo and Torre river system, that was located in the Grado area. The actual Grado coastline and the Mula di Muggia bank are the main wrecks of this ancient morphology (Brambati, 1987).
During the last century many different human actions have been carried out here, in order to limit the erosion of the coastline and to keep active the channel ports. These have been, in the studied area, drainage banks strengthened by several sea combs, perpendicular to the coastline. These human interventions have often been not useful to solve the erosion problems of the area and they sometimes caused some considerable erosion and sedimentation problems along the waterfront (Brambati, cit.).
Using GIS and Remote sensing techniques we tried to delineate the evolution of the coastline in the studied area. The main problem we had to face has been the making uniform some different mapping sources. These were in fact ancient and modern paper maps, digital vector maps and satellite imagery.

Source of Data
The used database is:

  1. The corographic paper from the Franceschino land register, 1830 (scale 1:28.800).
  2. The IGM (Istituto Geografico Militare) tablets of the Italian Official Map, 1949 (scale 1:25.000).
  3. The Numeric Regional Digital Vector map (CRN) of the Friuli - Venezia Giulia Region, 1988/89, (scale 1:25.000).
  4. An Aster sensor, satellite Terra, image, dated September 2001 (Spatial resolution: 15 meters in the VNIR, 30 meters in the SWIR, 60 meters in the TIR; spectral resolution: 14 bands) (fig.1).
  5. All vector and raster data elaboration and mapping has been made with Erdas Imagine 8.6 and Arc Gis 8.2.

Figure1. The Grado lagoon and the mouth of the Isonzo river (Aster sensor, satellite Terra, image, dated September 2001)

Methodology
A flow chart of the adopted method is shown in Fig. 2.

Figure 2. Flow chart of the used methodology

a. Acquisition by scanner of the ancient and modern paper maps (the 1830 and the 1949 one).

b. Extraction from the CRN digital map of the following vector layers: main roads, built up areas, coastline and major rivers.

c. Rectification of all the digital maps and the Aster sensor scene (Projection type: Transverse Mercator; spheroid name: International 1909; datum: Rome 1940).
As regard to the technical data about the rectification process:

d. Overlaying of the CRN extracted vector layers on all the rectified maps, in order to verify the precision of the rectification. Fig. 3a shows the Franceschino Land register map with the CRN layer overlaid on it. Fig. 3b shows a particular of the rectification of the 1949 IGM map. The figure shows the island of Grado. The overlaid CRN layer (in red) shows the later construction of the Schiusa island (built in the 1957).
Figure 3a. The rectified Franceschino land register map with the 1988 CRN layer overlaid on it (in red).
Figure 3b. Particular of the rectification of the 1949 IGM map with the 1988 CRN layer overlaid on it (in red).

e. Digital drawing of the dry land areas in 1830 and 1949, using as reference the old rectified paper maps. Fig. 4 shows an example of the digitization process of three simple classes (Dry land, Built up areas, Sand), in the 1830 historical map.

Figure 4. Drawn categories in the rectified 1830 map.

f. Evaluation of the old map dry land areas in square meters. We rasterized with a 15 meters grids the drawn dry land areas respectively in the 1830 and 1949 maps. From the pixel numbers we calculated the dry land area in square meters (15m x 15 m x pixel number).

g. Classification of the Aster 2001 satellite image in 5 simple classes (Built up areas/Bare surfaces, Shoal surfaces, Vegetation/Cultivated areas, Water, Mula di Muggia bank). Like in the previous point we calculated the extension (square meters) of the dry land areas (all the classes minus Water and Mula di Muggia bank). Fig. 5 shows the classified Aster sensor image.

h. Digital drawing of the sandy coastline in the 1830 and 1949 maps, in order to locate the beach areas in these two periods. Digital drawing of the 1830 lagoon channels. Overlay of all these layers on the Aster sensor image (2001).

Figure 5.The studied area classified in 5 different classes. This map has been elaborated from the Aster sensor image

Results and Conclusions
We finally compared the dry land extensions in the three analyzed years (1830, 1949, 2001) for the studied area. Fig.6 shows the histogram of the dry land areas, whose extension has been transformed in squares Km.

Figure 6. Histogram of the dry land areas (Km2) for the studied area.

Observing the histogram is clearly evident the growing trend in the observed period. The major growth is visible between 1830 and 1949 (1,5 km2). Instead the extension of the dry land areas between 1949 and 2001 is almost the same.
As regard to the 1830 - 1949 growth, it is interesting to notice that the drainage executed in the marshes located in the north of Grado municipality at the beginning of last century seems to have been compensated by the erosion along the coastline Westside of the Isonzo river mouth. This can be clearly seen in fig.7a and fig.7b. This result is also confirmed by a previous study (Brambati, cit.) that puts in evidence the main erosion direction in this Adriatico Sea area (Fig.8).

Figure 7a. Drawn 1830 dry land areas overlaid on the Aster sensor image Figure 7b. Drawn 1949 dry land areas overlaid on the Aster sensor image

Figure 8. Principal transport, erosion and deposit processes along the coastline from Grado Island to Isonzo river mouth (source: Brambati, 1987).

From the Brambati figure is evident that the whole coastline of Grado is potentially interested by strong erosion processes. In fact overlaying the drawn 1830 and 1949 sandy coastline areas on the Aster sensor image (fig.9), is clearly visible a strong retrogradation of the beach during the observed period. While the beach behind the Mula di Muggia bank seems to have been less affected by the erosion process, the tourist APT beach, more exposed to the Bora winds that blows more frequently in winter time, has had a stronger erosion.
Finally in fig. 10 it is shown the overlaying of the 1830 channel network on the Aster sensor image. Observing the picture is clearly evident where an ancient channel called "La Sacca" has been filled in 1938 (De Grassi, 1988). This human intervention may have contributed to destabilize the ecological balance of the water environment.

Figure 9. Drawn 1830 and 1949 sandy beaches overlaid on the Aster sensor image.
Figure 10. Drawn 1830 lagoon channels overlaid on the Aster sensor image.

Concluding, it seems to us that the proposed methodology is quite useful in order to check the main evolution trend of a territory. In fact we delineated some human made spatial changes. These ones seem to be in accordance with the model proposed in 1987 by Brambati.

References