APPLICATION FIELDS AND DATA PREPARATION FOR THE WEB-BASED COASTAL GIS OF PUCK LAGOON - SOUTHERN BALTIC

Jacek Urbanski, Katarzyna Bradtke

University of Gdansk, Institute of Oceanography (PL)

Introduction
Advance in web-based GIS solutions for oceanography is closely related with the possibilities of internet map servers products. The iteration process to make use of benefits of the latest achievements in internet technology is observed. The applications used to view, quarry and download data were the ones to be developed first (Valavanis, 2002). They were followed by projects designed mainly for sensitive marine areas. Using maps overlay in different scales they allow to assess habitats and disseminate data. Apart from using typical layers such as coastline and administrative borders these systems use bathymetric and biological maps, usually maps of habitats and environmental sensitivity indexes (MRGIS, 2003). The systems also fulfill education purposes by linking the shapefile features with multi-media still photos, videos, figures or descriptions. In the last few years research in regional oceanography has become more interdisciplinary than it was earlier and this process is increasing. Interdisciplinary research requires the assimilation of data from many different sources or themes. The GIS systems in geology, chemical, physical and biological oceanography have proved to be a useful tool for interdisciplinary analysis (Wright, Bartlett (ed.) 2000, Wright (ed.) 2002). It is only a matter of time when the web-based applications will start to play the same role. Easy access may cause that in the near future these systems will play an important role in interdisciplinary research as well as in other types of research that use physical background. Such layers as temperature or salinity are undoubtedly very useful in most projects. Variables of these layers change in time hence creating theme maps with such data causes the problem how to find a compromise between temporary accurateness and all disadvantages of various statistics. Because of rapid changes in technology what seems more important than informatic aspects are the methods used to prepare the layers and the search for widespread usage of the systems.
The aim of the project was to create a web based GIS system that could not only be a useful tool in interdisciplinary research but could also be of use for those in charge of management and administration of the coastline. This system was set for management and research of the Puck Lagoon, a small but essential part of the Polish coast of South Baltic.

Figure 1. The Puck Lagoon - the area of the project and its bathymetry map. The blue rectangle shows the area of large scale analyses.

The Puck Lagoon (Figure 1) is an essential area of research and management. It consists of two main parts divided by a shallow barrier: a shallow part with a rich environment of seagrass meadows and a deeper part where local upwelling and influence of fresh water play an important role in hydrology. In the winter, the shallow part of the Lagoon is covered with ice. In comparison with the coast of the open Baltic, this complex environment has a great biological diversity. The Puck Lagoon is designated to be a green star spot in the European biodiversity research program. Several species, like the common seal or Baltic sturgeon are currently being reintroduced. The Lagoon is also a place of many conflicts between man and environment. Strong erosion is a threat to coastal infrastructures in some parts of the Lagoon. Nearly half a million people live in vicinity of the Lagoon with the coasts flooded by tourists every summer. For several research institutes as well as for the University of Gdansk the Puck Lagoon is a place of diverse research projects.

Designing the system
The first step was to name the purposes for the GIS system. Two main questions for which an answer will have to be sought have been set. Who is going to use the system? What are the main purposes of these users? It has been assumed that two main groups of people are going to use the system. A group of researchers using the system for interdisciplinary analysis that will also find the system helpful in planning experimental research and a wide group of professionals in charge of coastal protection, coastal management and emergency response. As the system will be mainly used as background for interdisciplinary research, users will mostly need basic layers describing the environment such as bathymetry, sediments, temperature and salinity. Aerial photos of the coastline and shallows may be helpful in research planning. Layers showing borders of protected areas, rocks on the seabed, dumpsites may also be of use. The main difference between the work of researchers and those in charge of coastal management is the scale of their work. Most research takes the whole Lagoon into account while management of the coastline requires analysis on a local scale (part of the coastline). Processes of erosion and accumulation, the threat of flooding and spreading of pollution are typical matters of analysis. To bring out the analyses also layers describing the infrastructure of the coastline: roads, buildings, rivers and aerial photos are needed. The analysis of the requirements of potential users allows to decide which layers will be needed (Table 1).

Lagoon scale (research)
datasets GIS format comments
coastline line  
bathymetry

polygon
line

 created from raster bathymetry map
contours for overlay in visualization
salinity polygon average salinity on surface and bottom
temperature polygon temperature scenarios on surface
sediments polygon  
borders polygons  
aerial photo linked pictures  
Local scale (management)
datasets GIS format comments
bathymetry polygon  
change of coastline polygon erosion and accumulation in the last 100 years
flooding polygon areas flooded at given sea level
discharge polygon discharge from river mouth and sewage pipes
roads line  
rivers line  
buildings polygon  
maps and aerial photo raster old maps and registered aerial photos as background for analysis

Table 1. Basic layers of the system

Due to the fact that web-based GIS was created using HTML viewer of Arc IMS 4.0 software, all layers had to be designed as vector layers, raster layers could only be used as background.

Layer creation
The quality of the system depends on the functionality of the layers and this creates space for possible applications. One of the main problems in designing a system which uses IMS Esri is the limitation to use only vector maps for analyses. An essential part of the maps carries the description of continuous fields like depth, temperature, salinity, hence the problem of changing raster data to vector data is bound to occur. In the discussed system, raster maps for ex. a map of bathymetry had been reclassified to classes of 1m depth range and vectorised to polygons. The map of bathymetry was also created as a contour vector additional layer useful when overlaying it on any map of the system for better visualisation. The layers describing salinity and temperature create a background for many analyses. In physical oceanography salinity maps are usually created for specified levels. From the point of view of biological oceanography, however, the distribution of salinity on sea bottom, which is a habitat for many organisms, is much more important.

Figure 2. Map of salinity distribution on the bottom made using ordinary cokriging. The primary data was average salinity at the bottom. The secondary data were randomly selected bathymetry points and points of salinity estimated at 5, 10, 20m levels in intersection zones with the bottom.

Maps of distribution of continuous variables are created by interpolating point data. A map of salinity distribution on the surface (data from the last 20 years) has been made by interpolating average salinity measured on 25 stations. The fact that salinity is correlated with depth must be kept in mind when interpolating data from places where depth varies. Geostatistical methods, out of which ordinary kriging is most commonly used, give various solutions useful in creating layers of GIS systems. Cokriging allows to interpolate one variable (called primary variable) with the help of secondary variables which are correlated with the primary variable. This method gives especially good results when the number of secondary data is greater than primary data. Randomly selected bathymetry points and points of salinity estimated at 5,10,20m levels in intersection zones with the bottom had been used to create the map of salinity distribution on sea bottom. Data sets and the map are presented at Figure 2. Designing the maps describing the distribution of temperature on sea surface creates a lot of problems in areas of seasonal variability. Strong temporal and spatial diversity of the temperature of sea surface can be observed in the analysed area in spring and summer time. For these times maps of temperature obtained from AVHRR images had been used to create typical temperature scenarios on sea surface (distribution of cold and warm waters). A PCA analysis of 26 distinct in time maps of temperature had been brought out. This allows to present a great part of the variability with a limited number of five different maps of possible scenarios of temperature covering 80% of all possible situations. This method proves to be useful in detecting upwellings or temporal whirls which may be connected with biological phenomena.

Figure 3. Layer of erosion and accumulation. Overlay registered map (1910), aerial photo (1948) and present coastline (red line) give polygons of erosion or accumulation for particular periods (1910-1947, 1947-1997).

One of the basic functions in vector GIS systems is the possibility to overlay layers in order to create a new layer of polygons (union function). This new layer is created using a logical statement with variables having features of several layers and having inherited all the attributes of parents' layers. Maps of suitability and vulnerability, which are often a final product in GIS analysis, are created in this way. In order to create this kind of maps in IMS system, a special analytical layer (a result of union operation on a sequence of layers) has to be designed. In this system such a layer has been prepared using sediment, depth and salinity of sea bottom layers respectively.
Maps designed in a local scale have been prepared to solve set problems. These have been prepared for the area marked as a blue rectangle in Figure 1. Only areas important in coastal management are to be taken into account. The creation and analysis of maps of erosion and accumulation of the coast has been discussed by many. The analysis of changes in coastline has been brought out using a 1910 topographic map as well as 1948 and present aerial photos. The material was registered to UTM33N projection. In the area of interest the same roads, canals and buildings are present on all three sources. This fact made it easier than expected to register the material. Areas of accumulation and erosion are shown as polygons by comparing the coastline for the time periods: 1910-1948, 1948- present day. The next step was to overlay these polygons using union function. A result of this was a layer showing the process of accumulation and erosion over the last 100 years (Figure 3). Layer showing which areas are threatened with flooding at certain sea levels has been prepared. The probability of flooding was determined by analysing a dtm map with a known RMS of elevation (Urbanski, 2001). Survey works were a base for a detailed dtm map with the dune ridge well mapped. Next, the RMS of elevation was calculated. The probability of flooding was determined by assuming normal distribution of RMS. Areas with probability greater than 75% were taken as flooded. For each sea level a separate layer has been made. Then all layers were overlaid. As a result one layer showing flooding for different sea levels had been created. A separate group of layers are layers of water discharge from rivers. These layers are a result of experimental research carried out in various conditions. Water discharge is related with wind direction and the shape of the river mouth. The layers of river water concentration (in %) are shown as polygons. In Figure 4 water discharge in conditions of southern wind of 5-6m/s and shallow bar in river mouth is shown.


Figure 4 Map of water discharge concentration shown as polygons in conditions of southern wind of 5-6m/s and shallow bar in river mouth.

Application fields
Technical abilities of the system depend on the kind of the layers available and analytical abilities of the system itself. This system has been designed by Esri's Internet Map Server 4.0/4.1 for use with an HTML viewer. All layers of the system are designed in UTM33N projection, hence it was important to visualise the pointer's position in the geographic coordinates. The system was also prepared to give attributes of all visible layers of the objects identified with the function of identify. This led to gaining information from all layers in a point of given coordinates. Data from these layers may be treated as additional information in experimental research. The system might be also helpful in planning the position of a sampling station, particularly if the station is to be a base for research combining several factors (sediments, depth, salinity, distance from coast). Stating which areas fulfil certain conditions when designing maps of suitability and vulnerability is a similar case. This aim may be obtained with the use of quarry function on the layer constructed by using union operation on factor maps. In Figure 5 the result of the search for a specified area using the method presented above can be seen. The area was to meet certain conditions: depth not greater than 15 m, sea bottom with sandy sediments and average salinity at bottom greater than 7.6 psu. With the use of quarry with statement: SEDIMENTS = 4 AND DEPTH = 15 AND SALINITY >7.6 an area has been found. Such analysis gives a wide range of possibilities in interdisciplinary research, which can be widened still by using traditional methods as buffer and distance.

Figure 5 A map obtained with the use of quarry with statement: SEDIMENTS = 4 AND DEPTH =15 AND SALINITY>7.6

Application possibilities of coastal management have been defined clearer than analytical flexibility of interdisciplinary research. The analysis of erosion and accumulation is brought out by overlaying photos or map of the coast (map from 1910, photos from 1948 and 1997) and a prepared layer of erosion and accumulation. The use of distance method allows to analyse the changes in coastal processes and their tendencies. In the area of analysis the maximal rates are 2m per year for erosion and 2.3m per year for accumulation (Figure 3). One of the causes of vast changes in this area of the coast was the deletion of one river mouth in the first half of the twentieth century. The system allows for profound observation of the changes in the river mouth throughout the last 100 years.
Another important usage of the system can be the analyses of flooding. Layers of polygons covering the areas flooded with the sea level given (and greater) are used. This allows for a versatile analysis with the use of one layer only (Figure 6). Finding the flooded areas as well as places in the dune line where water overflow is bound to occur is possible. The analysis can be useful to examine not only the results of storm searches but also the effects of gradual sea level rises. Figure 6 shows that in the area of analysis the low areas behind the dune ridge can be flooded with the sea level greater than 560cm.


Figure 6.The analyses of flooding show which part of the coastal zone will be flooded at specified sea level

Yet another usage for the system is solving the problem to what extent pollution discharge has an effect on coastal environment. Buffering the polygon with defined concentration of pollution gives the answer to the question which elements of the environment may suffer most. The system may be used in emergency situations concerning toxic pollution and oil spills, owing to the fact that the layer of pollution can be modified in real time if need be.

Conclusions
Simple analytical abilities and the limitation to analyse vector layers only is not an obstacle when designing internet GIS for oceanographers and those in charge of coastal management. The layers can be useful in a wide range of research. The ability to create maps of suitability and vulnerability of the sea environment and coastline can be an example. Raster layers being either the result of interpolating point data (results of survey works) or an output of numerical models can be converted to vector layers of polygons. This allows to use these layers in analyses. The system can be created on two levels: for experimental research and for coastal management. These levels differ as far as scale and layer type is concerned. The layers may be modified when needed, hence systems working in real time can be created. Tests show that for those who do understand GIS systems and know how to use them, the system was not a problem. All this allows to draw the conclusion that knowledge of GIS systems should be widened among oceanographers and professionals in charge of the coastline.

References