Australian Institute of Marine Science
P.M.B. No. 3, Townsville M.C., Townsville 4810, Queensland, Australia
with the support of
The IBM International Foundation
The coastal oceanography group at the Australian Institute of Marine Science (AIMS) is currently undertaking research in an on-going study entitled "Coral Reefs and Mangroves: Modelling and Management" (CRAMMM). The project has a multi-disciplinary approach with field and simulation work under way into a number of processes in the coastal marine environment, including fish recruitment to coral reefs, sediment transport, large and small scale hydrodynamic modelling, and pollutant transport. This work is being undertaken in a variety of tropical ecosystems and locations, including the Great Barrier Reef and coastal mangrove swamps in North-East Australia, and several locations in South-East Asia including Vietnam, China, Malaysia and Thailand. Developing an understanding of coastal processes for both scientists and managers is critical if sustainable use of the coastal environment is to be maintained for future generations.
The primary visualisation tool of the project is IBM's Visualisation Data Explorer (DX) software. DX tools have been developed to view and animate both simulation results and field data. These tools allow observations and predictions to be displayed simultaneously for calibration of numerical models. A variety of field data have been visualised, including oceanographic, fisheries and coral biology, and pollutant chemistry data. The visualisation tools have in many cases provided the first opportunity for data sets to be viewed in their entirety, and have made an understanding of the patterns in the data possible on a scale never before achieved. In addition, the visualisation techniques enable full communication of scientific results to managers and non-scientists alike. Such visualisation technology is becoming an invaluable asset in the future effective management of the marine environment.
Developments in visualisation technology are changing the way science is undertaken in many areas of research. Modern computer hardware and software offers an unprecedented and rapidly increasing capacity to store, manipulate and display large volumes of data. The power and flexibility of visualisation tools are seeing a shift in emphasis from communication of results to a more active role in the research process. Additionally, powerful visualisation tools have the capacity to bridge the gap between scientists and non-scientists, making technical results much more accessible to all members of the community. This has many possibilities for use by environmental managers responsible for the long-term viability of renewable coastal marine resources.
The problems encountered in applying modern visualisation technology to the areas of coastal oceanography, fisheries and coral biology, and pollutant transport, are by no means unique. However the range and diversity of the difficulties encountered are likely to find parallels in almost any field of modern research, and as such the results presented here will have relevance to a wide range of subjects.
The data we are typically dealing with is characterised by
This applies to both field data and the results of simulations. The variables which are frequently measured include wind and current speed and direction, water salinity and temperature, suspended sediment concentration (SSC), particle and floc size distribution, pollutant and trace contaminant concentrations, proportion of organic to inorganic material, calcareous to terrigenous sediment, and so on. Distance scales range from less than 1 metre to 100 kilometres, and time scales from 1 second to 1 year.
As a consequence, modern visualisation technology is extremely appropriate to apply to this research area. This technology can have an impact in a number of ways. Researchers have infinite choice in the way the data is to be displayed. Applications can be designed to be interactive to some extent to allow the viewer to change the presentation of the data during the visualisation process. This results in visualisation methods which are the most appropriate, effective and intuitive for the subject matter and allow the best possible understanding of the processes which are described by the data (King et al., 1996). With better understanding comes increased predictive capability and more directed future research.
Visualisations can be tailored for a particular audience to ensure better communication of results to all interested parties. A complex, multi-variable visualisation, with custom-designed glyphs representing various experimental parameters may be easily understood by someone who is familiar with the data and the situation described, but will be meaningless to others without extensive interpretation. Both environmental resource managers and members of the public are groups who have a vested interest in the health of their local coastal area, but who may not have the scientific background required to fully comprehend technical scientific results obtained in that area. 3-dimensional visualisation and animation of results increase the capability of the viewer to link the scientist's model of the processes to the situation in the real world.
Figure 1. Schematic describing the Coral Reefs and Mangroves: Modeling and Management (CRAMMM) project.
The CRAMMM project, sponsored by both AIMS and the IBM International Foundation, is focussed on tropical ecosystems including coral reefs and mangrove habitats (Figure 1). These environments play an important economic role for many countries, including several in the South East Asian region which are currently undergoing rapid population and industrial growth. Increasing pressure on these sensitive habitats, coupled with a lack of environmental awareness, have resulted in widespread degradation at an alarming rate. The current project seeks to address these issues in two ways. Firstly, by improving the capability of predictive numerical models to simulate processes in these typically very complex systems, through fundamental research and the integration of existing multi-disciplinary data. Secondly, by developing visualisation software which can bridge the gap between scientists and research managers and place powerful predictive tools in the hands of the managers who will ultimately be responsible for the success or failure of those ecosystems.
The project is currently in its third year. The first year focussed on the development of visualisation tools and combining the results of the oceanographic and fisheries and coral biology research (Wolanski, 1995). The second year involved further refinement of the visualisation and modelling technology and an effort to transfer this capability to researchers working in various regions of South East Asia, including China, Malaysia, Vietnam and Thailand (Wolanski et al., 1996). The objectives for the third year are to package the modelling and visualisation software into a useful predictive tool for resource managers. The project has been extremely successful so far, resulting in more than 10 publications and generating wide interest from many areas in the community. For more information see the CRAMMM Project Home Page at http://ibm590.aims.gov.au.
Project hardware includes one IBM RS/6000 model 590 server configured with 1 GB RAM and 12 GB disk space, with a further 8 GB of on-line storage from a high performance RAID disk array. This server acts as a combined modelling/visualisation engine catering to a number of smaller workstations. RS/6000 models 390 and 39H, each with 256 MB RAM and 4 GB of disk space are used as secondary modelling/visualisation workstations. The model 590 and 390 servers are connected by a FDDI link providing high-speed data transfers, and in addition the model 390 workstation is equipped with a GXT-1000 graphics accelerator. Two RS/6000 model 25T power PC's, an RS/6000 X-station model 160 and two Aptiva Pentium 133MHz PC's complete the system. Each of the workstations are connected to each other and to the rest of the AIMS network via thinwire ethernet adaptors.
The software complement includes AIX v3.2.5 as the main operating system for the workstations, and Windows '95 for the PC's. AIX Fortran v3.2 is used for running all numerical models and data processing software. Data Visualisation Explorer v3.1 is the main data visualisation platform.
In this section we present some examples of how Data Explorer has been used to represent various types of field data during the project (Galloway et al., 1995).
Visualisation of field data has been an important component of the project. AIMS has been undertaking research on tropical ecosystems for almost 30 years and has accumulated an invaluable data set on a wide variety of processes. In some cases limitations of data visualisation technology has prevented full use being made of this data. One of the particular areas of interest for this project is oceanographic and behavioural influences in the dispersal of coral and fish larvae. In one experiment, light traps were deployed at 17 sites around a single reef, and fish larvae catches were recorded every night for two weeks (Wolanski and Doherty, 1995). Each catch was analysed and the count from each of approximately 2000 distinct species was recorded. Data of this volume clearly makes visualisation by traditional means extremely difficult.
This field data set played a very important role in the coral reef component of the project. One of the most important questions in determining the level of protection accorded to different coral reefs is to what extent is each reef responsible for acting as a source of coral and fish larvae, both for replenishment of its own stocks and those of other reefs in the local area and further downstream. The data described above offers a unique opportunity for insight into this recruitment process at very small scale and for a wide variety of species. Because of this, it was deemed necessary to better understand the data and hence a visualisation solution was sought.
The basic technique is very simple. The primary component of the visualisation is a surface representing the fine-scale bathymetry, with exaggerated variation in the depth. Measurement sites are represented by pins, the heads of which are the anchor points for some custom designed glyphs. A vertical bar created with the Glyph and Tube module represents the catch for a single species at that site (Figure 2). In addition, a coloured disk could be added at the base of the bar represents the proportion of the total catch at that site over the experiment duration caught at that day. The DX network was designed to accommodate the requirements of each user, and data for individual fish species or families comprising a number of species could be viewed. The data could be animated with the use of the Sequencer. An occasional criticism of this technique by casual viewers was that it was far too complex, and essentially this is true for a casual viewer. However the visualisation was never intended for the casual viewer and was instead aimed at the researchers who gathered the data, and were very familiar with what the data represented. The application was developed in close consultation with these scientists, who, when presented with unlimited freedom in regard to how the data was presented, opted for a method which was certainly complex; however this allowed them to scan the data rapidly and detect patterns in the measurements as easily as possible. The result was certainly a success from their point of view, but this experience provides a very instructive lesson: that no matter how powerful a visualisation platform is, each display must still be carefully tailored to the audience. The best results are possible only when this is done. Conversely, this demonstrates that not every visualisation is suitable for every audience.
Figure 2. Fish larvae light trap catch around Bowden Reef, Great Barrier Reef, Australia. Data glyphs are superimposed on 50m resolution topographic surface of the bathymetry of Bowden Reef.
A further development of this idea presents an even more complex example for most casual viewers. In this example, the requirement was to display a much larger number of variables with a technique similar to the one described previously. Long-term oceanographic data was obtained from moored buoys at seven sites within the Gulf of Thailand. Wind and current speed and direction, and salinity and temperature at four different depths were measured at each buoy location every hour over 16 months . Obtaining oceanographic data on this scale is extremely difficult from a logistical point of view; instruments can occasionally malfunction or fail, and this dataset had no shortage of gaps. One of the largest jobs in preparing this display was simply the pre-processing of the data to fix all the measurements to a constant timestep, at fixed positions, and replace all spurious and missing readings with some appropriate data value indicating a blank. However once this was complete importing the data into DX and animating it was straightforward.
The custom glyph idea used in the previous visualisation was extended to represent the larger number of variables at each measurement site. Measurement sites were again indicated by pins fixed to the topography. Current speed and direction was represented by an appropriately scaled arrow, distinctly coloured to stand out from the other display elements. Wind speed and direction was indicated by a "windsock" created with the Glyph module which was thought to be slightly more intuitive than an arrow. Salinity or temperature at the four different depths were represented as appropriately coloured spheres fixed at different heights on a separate pin offset from the measurement site. Finally, a coloured disk representing surface salinity surrounded the pin. The visualisation application was configured to be flexible enough to allow the user to switch each display element on or off individually, and again the data was animated to show variations with time. To show longer-term variations the dataset was averaged over each month outside Data Explorer, and this new monthly dataset could be Imported as another user option (Figure 3.). This visualisation resulted in important insights into the seasonal variations in current conditions in the gulf.
Figure 3. Seawatch data display in the Gulf of Thailand, monthly averages from April 1993 to August 1994. Multivariable glyphs are defined as follows:
In June 1996 a researcher from Thailand visited AIMS with some extensive data on the levels of dissolved and dispersed petroleum hydrocarbons (DDPH) in the Gulf of Thailand. It was a simple matter to add this data to the existing visualisation, representing the DDPH levels as bars whose height and colour depended on the measurement at that site. While this addition added elements to an already complex display, the visualisation provided important insights into the seasonal variations in pollution levels in the gulf. The data show widespread low-level pollution with much higher levels occasionally observed, generally located in the inner gulf area where industrial activity is greatest (display not shown). Additionally a seasonal variation was detected, which resulted from yearly variations in the circulation regime and hence the mixing rate and residence times of the gulf (Wattayakorn et al. 1996).
Another experiment where Data Explorer played a vital role in visualisation was the 1982 study of the eddy behind Rattray Island ( Wolanski et al., 1984 ). Rattray Island is located in the Whitsunday Group off north-eastern Australia. The island is small, approximately 1000 by 300 metres, and located in an area of shallow water with strong tidal currents. The 1982 experiment involved deploying up to 26 current meters simultaneously to obtain an accurate picture of the eddy which forms in the lee of the island at tidal frequency. Eddies in shallow water can be an important forcing in coral reef dynamics, as they cause upwelling of sediment and nutrients from the sea floor which can strongly affect coral and fish living on the reef.
Previous visualisation efforts with this data had been limited to static plots of the measured current field at a few times . Using Data Explorer it was possible to Import all the current meter data and provide animated plots of the field data, with current speed and direction represented by arrow glyphs. In addition, current speeds between the measurement sites was indicated by a coloured field overlaid on the field measurements. This coloured field was calculated by interpolating from the field data measurements using the Regrid module. Interpolated currents at the grid positions were also shown as arrow glyphs, coloured differently from the field measurements for clarity (Figure 4). It was also possible to display the vorticity associated with the circulation in this manner. The vorticity was calculated by collecting all the interpolated current grids from each timestep and combining them into a series object. This object was then stacked in the z-direction and the curl of the resulting 3D vector field was calculated using the DivCurl module. The z-component of the curl of each layer is then equivalent to the vorticity of the two-dimensional circulation in each layer and can be overlaid in the same manner as the current speed (vorticity animation not shown).
Figure 4. Current data from Rattray Island, north-eastern Australia. The black arrow glyphs represent the currents measured at each mooring site. The red arrow glyphs are interpolated currents calculated using the Regrid module and a regular grid. The colour field shows the current speed. The display in the lower right hand corner shows the tide height measured at the island. The horizontal field of view is approximately 6km.
The advantages of using Data Explorer, and in fact any modern visualisation technology, are obvious when compared to the more traditional techniques. These older methods are in many cases simply unable to display the large volume of data collected in these types of studies in any meaningful way. With the additional limitation of single colours and two-dimensional plots the capability of traditional methods to display multi-variable and multi-dimensional data is extremely limited. When the visualisation capability of DX is combined with its application building capabilities, the result is extremely powerful, intuitive visualisation tools which can be interactively adapted by the users and also easily modified to include new data or to modify the data representation. The animation capabilities allow temporal and seasonal variations to become observable, and the ability to Collect various object elements allows the data representation to be superimposed on bathymetry surfaces or images of charts to provide the very important geographical and topological cues.
Development of hydrodynamic models in a predictive capability is the other major component of this project. With extensive oceanographic field work over many years of study we have developed a strong understanding of the most dynamically important processes in both mangrove and coral reef zones. This work has typically been paralleled by modelling studies which seek to confirm the understanding gained from the field data by developing and testing models based on the critical processes in each domain.
Computer technology is an ever-present limiting factor in this type of work, and even more so before the use of powerful graphical workstations. Available processor speed and storage capacity determines the maximum number of discrete grid locations which can usefully be simulated. Given the geographical size of the domain this in turn affects the level of detail of the simulation. Without dedicated graphical displays results must be pre-processed before being displayed in normal black and white paper plots. This severely limits the understanding of the model results.
We have integrated several of the hydrodynamic models with DX. These models, exclusively written in Fortran, usually output an unformatted (binary) output file containing the simulation results. No change in the file format is necessary to allow these files to be read directly into DX, all that is required is a general or DX format header file which can be created by the model code at run time. The simplest display type for these models is identical to that used for the interpolated current data in the previous section, with currents at each grid point represented by arrow glyphs and an additional overlay representing current speed or vorticity.
An added complication is encountered when trying to visualise data from 3-dimensional models, where the currents are resolved on multiple layers in the water column. It is simplest to extract the currents from a single layer and display them in the same manner as is done for a 2-dimensional model, but there are other options which have been investigated. All the layers can be displayed at once with exaggerated scaling in the vertical direction to separate the glyphs vertically. It is necessary to have some kind of visual cue to indicate the relative height of each of the glyphs, and this can be done by colouring the glyphs depending upon their height.
It has been argued by some authors that a better way to represent current flow in an animated visualisation is by tracking simulated particles or streamlines (Rennie and Hamrick, 1992). Data Explorer permits this sort of visualisation through the use of the Streakline module. For the applications considered in this study, this method has limited use due to a number of factors. Firstly, it is sensible when simulating a tidally driven system to output the smallest dataset possible. This frequently means one or two tidal cycles, which can then be looped in a current animation to give an almost seamless join if the output begins and ends at the same state of the tide. It is not possible however to recycle current series in this manner using the Streakline module and for most applications the net movement of a streakline over one tidal cycle is very small. The second limitation is simply the processor time required to generate the streaklines. The greater the number of streaklines present in the animation, the better the resulting picture of the circulation. With a large domain of the order of 100 by 100 points, and strong horizontal current gradients, this quickly becomes an unmanageable computational effort.
What is more typically done is to run a simulation of this type as an external process and then re-import the results into DX. In this way a large number of particles can be used, and the simulated current data can be recycled as many times as necessary. In addition, it is reasonably straightforward to convert such a model into a more complex simulation by introducing variations in the behaviour of the simulated particles. For example, the particles can be regarded as pollutant particles, and a chemistry model can be linked to the transport model giving a much more detailed picture of the pollutants behaviour. Alternatively, the simulation could be considered as a sediment transport model, given the addition of erosion and deposition parameters.
One simulation using this type of model attempts to explain the variations observed in the numbers of fish larvae caught at different areas of the reef described in the previous section. This model postulates a cloud of fish larvae advected onto the reef by the net circulation, and examines the differences in results obtained by considering the fish larvae as passive neutrally buoyant tracers as opposed to actively swimming organisms (Figure 5). What becomes apparent from these two simulations is that the observed concentration of fish larvae in the lee of the reef cannot be explained without some active swimming component of the fish.
Figure 5. Results from a simulation of an incident cloud of fish larvae on Bowden Reef, Great Barrier Reef. Fish larvae are advected by the predicted tidally varying current field, and in addition will swim towards the reef at a constant rate whenever they are less than 3000m away. Predicted concentration in the lee of the net southerly current is similar to what is observed from the field data.
Some models lend themselves more to visualisation than others. One model for which a visualisation package has been developed is a 2-dimensional current, salinity and sediment model of Jiaojiang estuary, China. This model is distinct from other 2-dimensional models used in that it calculates variables on a grid in the x-z plane, that is, a "slice" down the centre of the river. Because the estuary is long and relatively narrow, with little cross-channel variations, this type of model provides a reasonable description of the dynamics. The model calculates salinity, current speed and direction, and suspended sediment concentration at each point in the x-z grid. Variable depth along the river is taken into account by "stretching" the layers between the surface and the bottom so that there is a constant number of layers at each horizontal position.
The visualisation package allows simultaneous display of all the calculated model parameters (Figure 6). The topmost frame shows the currents as arrow glyphs, overlaid on to a coloured field representing the salinity. The salinity field is banded with isolines separating the bands. Banding is used in preference to continous colour fields in some visualisations for two reasons. Firstly, when converting to a PC-based animation file in .FLC format as is commonly done, there is an inevitable reduction in the number of possible colours, and the wide range of colours used in a continuous colour map may result in uncontrolled banding of the colours which can change from frame to frame. This is of course dependent on the software used to create the animation files. The second reason is that it is sometimes difficult to determine actual data values at a point in a coloured field when a continuous colour map is used. While a banded field still does not permit actual data values to be determined it does permit knowledge of the data values within well-defined ranges and this can be very useful for certain applications. The middle frame shows the suspended sediment concentration (SSC), again banded for clarity. Note that the tidal height at each point in the estuary is indicated on both the current/salinity and the SSC display by the variations in the height of the field at each point. The tide height at a particular measurement site is also indicated on a line plot in the lower half of the image. The bottom frame shows the outline of the estuary, with a scale bar showing distance. The depth-averaged SSC, which is also calculated in the model, is Imported and calculated upon a grid using the Regrid module. This field is then coloured and superimposed upon the estuarine outline. The result is a very straightforward and clear display which nevertheless shows all the relevant parameters calculated by the model.
Figure 6. Jiaojiang Estuary hydrodynamic and sediment model visualisation. The data displayed, from top to bottom, are current (arrows) and salinity (coloured field); suspended sediment concentration (SSC, coloured field), tidal elevation near the mouth (line plot) with a marker showing the current time; and depth averaged SSC superimposed on an outline of the estuary. Tidal height at various points in the estuary are indicated by the thickness of the colour band in the SSC and salinity displays.
An extension of the modelling techniques discussed here is to display both the field and model data simultaneously. In this way, an intuitive, visual calibration of the model is possible. Other techniques are also possible, such as subtracting the measured values from the predicted values to get an animated visualisation of the error between the simulation and observations (Rennie and Hamrick, 1992). One difficulty which frequently occurs with this type of visualisation is that the start times and timestep of the field and model data will be different. This can fairly easily be coped with by using the Sequencer to select frames from one set of data, and then using the Compute module to calculate the corresponding frame from the other set.
The application development features of DX have been used fairly extensively during this project. The provision of interactors and control panels, and the capability to control access to various features of the user interface make configuring applications for non- DX experts a straightforward matter. Two applications which have been developed throughout the course of the project are worth mentioning.
The mod_view application is designed to permit common tools and interfaces to be used with output files from a variety of different models. The application utilises a DX format file containing information regarding each separate set of model code and the different regions each model is configured for. The file contains an entry for each domain and model which specifies the bathymetry file, where the output files can be located, grid size, and so on. The user can select the model type and then the domain, and is then presented with a list of output files from which the desired one can be selected. The list of available file names is generated by a filter as part of the FileList macro . Once the model file is chosen, the simulation data is Imported and the data is routed to the appropriate section for visualisation.
The application is configured to read data files from a variety of 2- and 3-dimensional hydrodynamic models, and sediment transport models. The user can choose to display data from a hydrodynamic model as arrow glyphs superimposed on a colour field representing current speed, vorticity or surface elevation. The user can position a probe in the field to extract a time series of current speed and water elevation at a single grid location. This time series is then plotted in a separate window using the Plot and StackPlot macros. For a 3-dimensional model, the entire current field can be viewed with the glyph layer indicated by its colour. Additionally, water surface elevation can be displayed in a 3- dimensional view as a Rubbersheet superimposed on a similar sheet representing the bathymetry. Alternatively, individual layers from the model output can be displayed in the same manner as for the 2-dimensional model outputs. Each of the display types can be animated with the use of the sequencer.
This application has proved very successful with several users within the group. The application allows model results to be viewed immediately upon completion of the model run, without any of the additional processing which was required before the use of this software, and without requiring a separate visualisation network to be written for each model. Adding a new model or domain to the application is a simple matter, requiring only that a general or DX format file exists for the model output and a new entry in the domain information file. Possible future developments include the use of external filters to remove the requirement for a separate header file for each model output file; the use of some type of "standard" binary file format, perhaps a widely used format such as NetCDF; and the addition of new visualisation tools as required.
A second application currently under development is called the Data Explorer Database or dxdb. This application, which includes external filters and custom macros, is intended to streamline the visualisation process for the wide variety of field data typically used in the project. The majority of data gathered is stored as ASCII files in a small number of different formats. Combining these data files for importing into DX can be a lengthy process. The principal component of the dxdb application is an external filter program written in AIX-Fortran which automatically combines different data files into a DX object. The filter reads an index file which lists the file names containing the data and any other relevant information about the data, for example the time the measurement was taken, the location in latitude and longitude, the parameters measured, and the units. The filter then combines this data into a DX object which is written to the standard output logical unit. When combined with the Import module in a Data Explorer network this filter program can be used to read in all the relevant data to the network.
Once the data has been imported, the user can select from a number of variables to view (Figure 7). Time series data can be animated, and a number of macros have been developed to manipulate the data.
Figure 7. Example of field data visualisation using the dxdb application. Current, wind and water temperature data taken around Lizard Island, north-eastern Australia, during November and December 1995. The base images is a scan of the standard nautical chart of the area. Image elements are defined as follows:
The contribution of Visualisation Data Explorer to the CRAMMM project cannot be underestimated. The visualisation tools which have been developed have made significant impacts in a number of ways. The capacity for display and manipulation of a variety of different types of field data has been significantly increased. Visualisation tools have greatly streamlined the modelling process by enabling immediate viewing of results. The animations produced have generated widespread interest in the project purely through their visual appeal. In addition they have allowed a wide range of people, from scientists to managers and the public to appreciate both the physical processes which we seek to describe and the management problems this project seeks to address. It is hoped that this work, with the help of the already strong visualisation component, will continue to generate interest in the environment and coastal ocean processes, and will have a significant impact on the future management of threatened coastal zones.
This study was supported by the Australian Institute of Marine Science and the IBM International Foundation.
Below are a selection of the new macros developed during the project.
 |
The Antialias macro is used to improve the appearance of images. It accepts an object and a camera and renders the object at two or more times the resolution specified by the camera. It then smooths the image by averageing over a number of pixels using the Reduce module. The third input controls the expansion factor. For example, if the factor is two, the final image will be the same size as specified in the camera. If the factor is four the image will first be rendered at four times the camera resolution and then reduced in resolution by a factor of two, resulting in a net increase factor of two. |
|
|
|
 |
This macro will band a colormap given a set of banding values. The Band module can be used to band data which can then be coloured based on any colormap, but a colorbar will still show the continous colormap. This module will band the colormap as well, based on the same set of values and banding parameter (remap = "high" or "low"). The colormap can be banded before or after coloring the field. This approach is more flexible than the technique of banding the colormap by hand using the colormap editor because if the colormap is data driven and the minimum and maximum of the dataset changes then the banding values set in that way will also change. (Note that this macro also requires the ArrayDouble macro). |
|
|
|
 |
This macro is extremely simple and basically acts as a pointer to the file_list filter. The filter is a C-shell script which returns a list of the files in the given path and satisfying the filename pattern as a DX object which can then be sent to a Selector to allow one of the filenames to be selected. |
|
|
|
 |
This macro creates a simple scale bar which can then be incorporated into a visualisation to give a distance reference. This macro was used to create the scale bar in the Jiaojiang Estuary model visualisation (Figure 6). The user can define the foreground colour, what text labels to plot and the text parameters, and the overall length of the scalebar. |
|
|
|
 |
This is an extremely simple but useful macro which merely calculates, given a series object, the maximum value allowed for the sequencer assuming a starting frame number of 0. |
|
|
|
 |
This macro accepts a 3D grid representing some measured parameter and a 2D grid representing the bathymetry. The macro scales the z-positions of the 3D grid between zero and the maximum given by the data value at each point of the 2D grid. This is useful for example for scaling the grid positions from a sigma-coordinate model to real coordinates. |
|
|
|
 |
This macro stacks two objects in the y-direction. The module is especially useful for arranging two or more plots created with the Plot module. The macro uses another simple macro called MoveToOrigin which simply moves an object to the origin. |
|
|
|
 |
This macro extracts a time series from a series object consisting of 2D fields. The macro accepts the series object as the input, as well as the required point (in the grid coordinates) and the origin in the grid coordinates (normally [1,1]). The macro outputs the time series in the form of a scalar array defined on a 1-D grid with positions representing the series positions from the original series object. The macro also returns the position of the cell in world coordinates. |
Comments to duncang@ibm590.aims.gov.au
Last updated: 2 September 1996
All contents copyright (C) 1996, AIMS.All rights reserved.