NEW TOOLS TO IMPROVE DESKTOP SURVEYS Pablo Vengoechea (Telemediciones S.A.), Jorge O. García (Telemediciones S.A.), Email: <pvengoechea@telemediciones.com> Telemediciones S.A. / Cra. 46 94-17 Bogotá D.C. Colombia. Abstract: Nautical charts do not have enough information about marine environments, specially at deep waters; therefore alternative methods to optimize marine routes are been demanded. Using the new release of the GEBCO grid, is it possible to simulate and to detail a submarine cable route with acceptable precision as to help survey engineers in the process to specify the safest marine interconnections between landing stations. The following sequence shows the new submarine cable system between San Andres Island and Tolu Colombia. The model takes information from a special GEBCO database and then using formats translators, dumps the information into a GIS based on Mapinfo. 3D images can be produced giving a clear definition of the marine path and enhancing the way of DTS is usually presented. Wide swaths can be considered and then using regression analysis is it possible to recommend the survey center main line. 1. INTRODUCTION In planning, design and implementation of a submarine cable, survey data constitute the most critical part of all processes and the future of cable itself. The results and decisions depend on this information. The vulnerability of the cable, the installation cost overruns and the risks that may occur regularly, can be minimized significantly if a survey is well done. To get the most out of the survey is important to plan the route in advance as much detail as possible, including all possible constraints in developing the project as political boundaries restrictions, physical constraints on the ocean floor, the presence of fishing areas or exploitation of hydrocarbons, among others. In this way it is possible to establish a route survey in which several issues were considered and avoided several risks to the project. This paper aims to show the advantages of Geographic Information Systems (GIS) in submarine cable route planning on the desk based on the useful information of the GEBCO bathymetry. With these tools, it is possible to integrate easily all variables and planning the safest route for the survey. Later with the survey data, analyze and specify the requirements for interconnecting the two land sites. 2. THE SAN ANDRES TOLU SUBMARINE CABLE The submarine cable between San Andres and Tolú for interconnecting the Colombian island with the mainland was planned from the bathymetry of the world of the "General Bathymetric Chart of the Oceans (GEBCO). The project comes as a state initiative to achieve better communications systems in the island and getting real broadband services. The cable will have a length of 800Km approximately. Given the need for a quick and effective implementation of the system, we compiled all the information of maritime boundaries and the approximate location of submarine cables in the area and took to the format of geographic information Copyright 2010 SubOptic Page 1 of 6
system to be analyzed together with data from bathymetry. The resulting product is a continuous array of bathymetry data for the entire world. Considering the characteristics of the sounding methods, the grid isn t have the same resolution in shallow waters as in depth waters. The GEBCO Grid is more suitable to the representation of bathymetry in deep water. Figure 1. San Andres Tolú area 2. ABOUT GEBCO The General Bathymetric Chart of the Oceans is made up of an International Group of Experts in ocean mapping. The GEBCO_08 Grid is a continuous terrain model for ocean and land [1] with a spatial resolution of 30 arc-seconds. The bathymetry part of GEBCO grid is a combination of selected survey data and gravimetric data obtained by satellite. The information comes from several contributors in the entire world. The GEBCO experts analyze its quality, before including it in to the general chart. Figure 2. GEBCO_08 grid Figure 3. GEBCO_08 Source Identifier Grid The GEBCO_08 Grid is stored in a standard netcdf File format using a onedimensional array of 2byte signed integer value. To import the data to our tool, we omit the header information witch consist of 612 bytes and read the grid as a generic binary format with 21600 rows, 43200 columns and a pixel size of 0.0083333333 degrees. The grid is projected as longitude/latitude WGS84. The GEBCO data comes with the source identifier grid (SID) to identify which cells in the GEBCO_ 08 Grid are based on bathymetric soundings and which cells contain predicted depth values. 3. TRADITIONAL DESKTOP PRE- SURVEY When planning a submarine route from the scratch, the usual procedure to follow is querying the available nautical charts to see the depth range in the trace of the preliminary route. With this information the designer have to estimate the possible types and lengths of both armored and unarmored cable that will be necessary in the entire route. Besides the approximate bathymetry of the nautical charts, the designer needs to consider the political boundaries information, the crossings with other cables and avoid passing over areas of elevated risks. While all this critical information lacks the needed precision, this will be the foundation of planning of the survey route. Copyright 2010 SubOptic Page 2 of 6
Once the survey route is defined from this information so poor, there is no guarantee that they don t need to inspect an area many times in order to find a better route. In the particular case of the submarine cable between San Andres and Tolu in Colombia, the planning started with the available data on nautical charts. The better bathymetry information that is possible to find in Colombian official maps is the nautical charts with 1:250000 scale for open sea areas or 1:25000 in areas near from the shore. The bathymetry data portion is these charts consist of depth values regularly spaced from 1 to 20Km in average. In this charts the contour lines are few, so the degree of uncertainty is large and the map information is not able to show the terrain variations of the depth of the sea. 4. GIS SOFTWARE TOOLS With the availability of the GEBCO data, planning a submarine cable on desktop using a Geographic Information System becomes more effective. The data can be imported as a raster grid and can be analyzed with different practical tools. The process of the route planning can be made iteratively. After the location of the land sites the search of the most viable route can start by means of a simple observation on the map, in this case is very useful to have the contours of bathymetry data over the other layers. The contours show easily the location of undersea hills, canyons or high slope depressions. The designer can construct a polyline trying to avoid the principal obstacles. Figure 5 Contour map from GEBCO grid Figure 4 1:250000 scale nautical chart from CIOH (Centro de investigaciones Oceanográficas e Hidrográficas) [2] In this first approach of the design, the resultant length of the route was 770Km, which is near to the length of the straight line joining the two points. The maximum slope can not be determined due to the length between alter courses and isobaths. In previous surveys and in the nautical charts there was found small rocks at 25 miles from Tolu, barely noted in the GEBCO model. The contour lines are obtained through a raster to vector function of the tool. The interval of values between contours can vary freely according to the needs of visualization, but the resolution of the grid imposes its own limit. The GEBCO_08 grid has a resolution of 30 arc seconds (around 1km), so it cannot show spatial features which size is bellow of this value. After the manual drawing over the map of a first version of the route, it is possible to do cross sections over different segments of the scene to read the depth variations. Copyright 2010 SubOptic Page 3 of 6
conference & convention The final survey route is the central line of the resulting region. Figure 6 Cross section More adjustments for the route can be made by means of 3D scene views and slope calculation. This option gives a clear view of the problem and helps to evaluate the better choice. Figure 7. 3D view San Andres - Tolu Figure 9. Obtaining the survey route plan. The region around the line represents the effective sonar area. The orange line is the axis of the region. Figure 8. Slope Grid When adding multiple layers to the map, like maritime boundaries, routes of existent cables, fishing areas or petroleum exploitation, the designer can analyze different options and calculate the impact of the different variables in the project. The preliminary red path in figure 5 is the calculated route for the project. With GIS tools, you can create a region around the line whose radius of separation will be a function of bathymetry values. This region represent the area covered by the sonar which depends on the depth of the ocean. Copyright 2010 SubOptic There are many differences between the results given by the nautical charts in the traditional pre-survey and the results derived from GEBCO data. With GIS tools and GEBCO data, the resultant route length was of 822Km. with a maximum slope of 72 degrees at Kp 186. We used a software tool based on MapInfo to do the analysis, but it can be made with different GIS software tools available on the market. The important thing is to import the grid as a raster map, having the possibility of reading the bathymetry data using profile tools or doing queries on grids by coordinates. 5. GEBCO DATA COMPARISON WITH SURVEY DATA The GEBCO data has a resolution of 0.008333 degrees or 1Km approximately. Page 4 of 6
This means that the minimum object size that can be discerned will be superior to 1Km. However, with this resolution the designer can see at least a regular variability on the ocean floor. To give an idea of the resolution of GEBCO data, we establish a particular case for comparison of the two sets of information. We use a sounding map from a submarine cable project in the Caribbean ocean with a scale of 1:100000. With this map we generate a digital terrain model of 100 meter resolution after the digitalization and interpolation of the bathymetry contours from the map. The used interpolation method is the triangulated irregular network model. A typical comparison method between two digital elevation models is to evaluate the maximum slope value. In a high resolution map, what you expect to see is a wide range of variation of the slope which minimum value is 0 degrees and the maximum value is 90. When analyzing the slope variation in the same area of the two sets of data, the GEBCO grid shows a minimum of 0.0104 and a maximum of 38.60, whereas the sounding data shows a minimum of 0.0008 and a maximum of 83.51. Survey Data GEBCO grid data Sounding - GEBCO difference Slope in GEBCO data Slope in sounding data Figure 11 Cross sections along a test route. Figure 10 GEBCO and Survey data 3D models. The figure 9 shows a 3D view of the two sets of data. The survey map shows a canyon with clear variations in its slope, whereas in the GEBCO grid the slope is smooth. However the presence of a canyon is seen on the GEBCO data, so the designer can take it into account in the planning stage. The figure 10 shows many variations in the slope along the test path over the sounding map, whereas the GEBCO data seems to follow the minimum values in slope of the sounding data. When comparing the cross section of bathymetry values, the GEBCO grid values are almost the average of the sounding data Copyright 2010 SubOptic Page 5 of 6
and all the variations over a path of 1Km length, are converted into one average value. 6. CONCLUSIONS The relatively recent release of the GEBCO grid bathymetry data gives a better alternative to start using the geographic information systems in submarine cable planning. The GIS software has shown a great advance in the last few years and many specialized tools can be easily applied in particular problems. Although the free bathymetry information from GEBCO don t have the required resolution for all cases, it is useful for a preliminary route planning and helps to minimize the problems during the survey process. Although the surveyor didn t consider these guidelines, the final comparison between the final center line and the GEBCO Grid, showed in specific sites a very good approximation, however in the shallow water crossing near San Andres, the surveyor didn t consider the alerts clearly visible in GEBCO model. 7. REFERENCES [1] http://www.gebco.net, The GEBCO Grid,http://www.gebco.net/data_and_prod ucts/gridded_bathymetry_data/documents/ gebco_08documentation.pdf, [2] CIOH (Centro de investigaciones Oceanográficas e Hidrográficas de Colombia) CIOH, Set of navigation maps, [3] "The GEBCO_08 Grid, version 20091120, http://www.gebco.net" [4] "The GEBCO_08 SID Grid, version 20091120, http://www.gebco.net" Copyright 2010 SubOptic Page 6 of 6