REMOTE SENSING COUPLED TO A DATABASE OF CATCHMENTS AND COASTAL ZONES

Stefanía G. Halldórsdóttir (1), Helgi Þorbergsson (2)

(1) National Energy Authority(NEA), Hydrological Service, Reykjavik (IS)
(2) The University of Iceland, Reykjavik (IS)

ABSTRACT
A great deal of coastal changes in Iceland are occurring in response to sediment transport and changes in glacier mass balance. In order to investigate the effects of sediment transport and glaciers on the coastline, research institutes in Iceland have started a project to develop a database to integrate remote sensed data, field data and spatial data on catchments and coastal areas. In the project, remote sensed data on coastal changes will be coupled to catchment data on land use, soil, vegetation, hydrology, sediment transport, glacier mass balance and freshwater biology. A Model of the data involved will be created and an assessment made on which data is relevant for the project. The research institutes have been working on integration of different scientific and spatial data into a common database. This has led to more cooperation between the institutes in the field of data management. The next step is now to expand the data collection, followed by the integration of remote sensed data into this common database, where a relationship will be established between all spatial and environmental data of different origin. The results will be presented by GIS, and it will be investigated how remote sensing data (aerial photos and satellite images) can be coupled to ground truth data in the best way so as to fulfil the aims of the project. Aerial photos and satellite images will be used to assess changes on the coastline. Database technology will be used to find a relationship between changes on the coastline and other environmental factors like sediment transport and changes in the extent of glaciers. The research will be carried out on one test area, the catchment of Ölfusá in south Iceland and the coastal area around the delta

INTRODUCTION
Research institutes in Iceland have developed a prototype of an online-database for environmental data (fig. 1). It was funded by The Icelandic Research Council and the research institutes and companies that were involved. The project integrates different scientific and spatial data from different research institutes, and makes the data accessible through web applications on the internet. The data represented in the prototype are freshwater data, oceanic data, land based data and geothermal data (Halldórsdóttir S.G. & Sigurðarson S.J. 2003).

Figure 1. Online environmental data.

This cooperation in the field of environmental data led to a further project which will also be funded by The Icelandic Research Council.
The objective of this new project is to develop a prototype of a research database for coastal zone management in Iceland, as well as methods for using remote sensing data coupled with a database on catchments and coastal areas. This involves remote sensing on coastal changes coupled with catchment data on land use, soil, vegetation, hydrology, sediment transport and freshwater biology.
In addition experiments will be conducted on how database technologies, as well as spatial analysis and data mining methods, can be used in the database to find relationships between environmental factors.

DATA OF WATERSHEDS AND COASTAL ZONES
The institutes involved will provide all sorts of data for the project. The participating research institutes and companies are: The Institute of Freshwater Fisheries, The Institute of Engineering (Division of Software Technology) at the University of Iceland, The Agricultural Research Institute of Iceland, The Hydrological Service and the Energy Resource Division of the National Energy Authority.
An important aspect of the project is that aerial photos and satellite images will be used to assess changes of the coastline. Database technology will be used to find a relationship between changes on the coastline and other environmental factors like sediment transport and changes in the extent of glaciers.
The research will be carried out on one test area, the catchment of Ölfusá in south Iceland and the coastal area around the delta (fig. 2).

Figure 2. The research area.

The database will be a holistic approach, as landuse data, hydrological data, marine data, biological data and coastal data will be brought together to one database. The catchment of Ölfusá is about 5678 km². Two glaciers are on the watershed, Hofsjökull and Langjökull.

DATA INCLUDED IN THE DATABASE
In Iceland, much of the water within the hydrological cycle, both groundwater and surface water, originates from glaciers, which cover approximately 10 percent of the land surface. Hence, most of the largest rivers in Iceland are glacially derived and transport great amount of sediment within their course from the glaciers towards the ocean (J. Harðardóttir & Á. Snorrason, in press). Research on sediment load in Icelandic rivers indicates a total flow of material into the sea from Iceland of the order of magnitude 50 millions of tons/year. Calculations of total suspended sediment load are made by correlation river discharge and sediment load, and the transported material is divided into different grain sizes, ranging from dissolved solids to gravel and stones (Haukur Tómasson, 1976). Many natural processes such as discharge, geology and glaciers influence the amount and the grain size of suspended sediments. Therefore, a database for coastal zone management for Iceland has to include abundant information on natural processes of the catchment, from glacier to ocean. In Iceland the glaciers have in postglacial time acted on the shore indirectly by yielding sediments for the rivers to transport to the coast. Glaciers and volcanoes act simultaneously through subglacial eruptions causing jökulhlaups which carry huge quantities of sediment to the coast. The sediment originates from three sources on land. These are:

1. Glaciers and glaciated areas.
2. Volcanoes.
3. Unglaciated areas.

The glacier-fed rivers take up moraine-like material from the basal layers of the ice. This mostly takes place in ice tunnels which the subglacial rivers melt and erode through the ice. At the glacier snout, this material is either deposited or carried further according to its grain size. The course gravel material forms bottom material and bed load. It moves relatively slowly and builds up "sandur" planes. The sandy silty material forms suspended load in the rivers and is carried with the running water directly to the sea. Volcanoes contribute directly to sediment load through ash eruptions. In Iceland there are two types of ash eruptions: sub areal and subglacial. The sub areal eruptions contribute through direct ashfall into the channels of the rivers on the drainage area. Subglacial eruptions sometimes create enormous jökulhlaups loaded with eruption material in the form of sandy silty ash, subsequently carried by the rivers. The third source of sediment delivery is soil erosion. Although quite high in Iceland, it is of minor importance compared to the other sources in this connection (Haukur Tómasson, 1985).

THE DATA MODEL
A model of the data involved will be created and an assessment made on which data is relevant for the database. The data will be collected from the institutes involved, followed by integration of the data into a common database, where a relationship will be established between spatial and environmental data of different origin.
The data model is an important task in the project; it will cover spatial-temporal, point and image data of different origin. The biological data from the watershed and from the delta will be coupled with hydrological data, both time series and point measurements. The model will also cover data from glaciers, sediment transport and marine data. Most of the data will be associated with a geographic object. Object modelling provides the ability to encapsulate a geographical object as a unit of spatial and attribute information along with methods that specify meaningful operations on that object (T. Helokunnas, 1993).
Since the structure of the data is very complex and heterogenous, it is important to structure the data set by using an object-oriented modelling technique.

DATABASE TECHNOLOGY
Natural systems can be considered as open, and reasons for environmental changes can be hard to estimate. It is necessary to include new database technology such as data mining and spatial analysis in environmental research. Good cooperation between the institutes involved is of great importance for good data management, and this approach is hopefully the beginning of an integrated research database in the field of watershed management and coastal zone management. In Iceland those two fields of research have to be coupled together due to natural conditions, and data management should be involved in the research process. The core of the project will be a central database, which will be the basis for further data management development and data integration. The data model in the project will be considered as a prototype for further integration between the research institutes.

DISCUSSION
In order to assure data integrity and to simplify the structure of diverse data, a comprehensive object-oriented data model is of great importance to the database for coastal zone management in Iceland.
By using object-oriented techniques, geographical, field and remote sensed data can be modelled, and it can lead to closer approximation to real world conditions. Sample data for example can be looked at as an object, with time and space as an attribute, and it should be possible to build time-series out of field data in the data set. The data model should also be expressive and it should be straightforward to extract semantic information from the data set.

REFERENCES