SVWebsite

by Milos Pelikan, Senior Analyst

The collation and creation of spatial information has seen phenomenal and renewed growth over the last few years.  This growth has been a consequence of the decrease in cost in a range of remote sensors (collection systems, including LIDAR), the need to update and improve existing information resources to tackle existing issues in natural resource management (NRM) and emerging spatial information requirements related to climate change, health and emergency management.

One of the emerging features of this new range of information is a tendency for there to be “stacks” or groups of related data – for example climate change variables represented spatially, time series from remote sensing and aquifer (groundwater) information.  This ”stacking” represents a new challenge for spatial technology as the end user is often as interested in the vertical or profile relationships between datasets as in the planar distribution of a particular dataset.  This issue of planar verses profile information; or rather the merging of both into a standard GIS environment, is the subject of this technical tip. In essence, the paradigm shift is to think of the profile as a fundamental dataset rather than an “output” ancillary graphic as presented in most GIS toolset.  In addition, and of equal importance, is addressing the geoprocessing challenge of mass production of profiles. Automation not only increase efficiency but enables the use of profiles to become standard practice rather than a speciality area of GIS.

The example

In this technical tip we use the example of recently developed aquifer (groundwater) information – sponsored by Southern Rural Water (SRW) – to illustrate key features of the new approach to the mass development of profiles. 

SRW recently commissioned a new generation of groundwater (aquifer mapping) information across southern Victoria. 

This information is seen as a key input to an increasing focus on groundwater resources and management as a result of protracted drought and the potential impacts of climate change to long-term sustainable water supply.

Key Issues

The core of our aquifer (groundwater) example addresses the following key issues:

• Aquifers are 3D in nature
• Aquifers represent a stack of related structural units
• The Vertical structure or relationships are critical to understanding and managing the resource
• The New “mapping” has been developed at much higher resolutions than previous studies
• Such stacked relationships are difficult in standard (planar) GIS

In addressing the above issues we will show that standard GIS technology coupled with novel geoprocessing techniques can be made to create information rich profile databases that represent aquifers, inherently 3D features.

Profiles of surfaces are not novel in themselves, in fact GIS by definition is an offshoot of that branch of mathematics called Graph Theory.  Any standard GIS represent information to the user as a two dimensional graph of X and Y coordinates that are variously describes as points, lines or polygons.  A third dimension “z” is most commonly associated with terrain or elevation, however may be other types of variables (eg. Humidity)

The crux of the new approach involves the use of geoprocessing to automate the development of a Database that is made up of many profiles rather than the standard profile graph that is a common ancillary tool to most mature GIS systems.

The aquifer information developed by SRW is represented by a set of raster surfaces (one per aquifer). Raster surfaces are a common and convenient way to represent complex or continuous variables across a geographic space (eg. Rainfall or Temperature).  While this is a well accepted approach by scientists and GIS professionals this data type can be confusing to many non-technical users of spatial information. 

Design Criteria

Two key design principles guided solution development in the groundwater example, these where:

• The solution must communicate these 3D or profile (“z”) relationships of the stacked surfaces
• The solution must be easy to access by resource managers and the community
  

What about 3D?

An initially obvious solution is to use 3D technologies as it is well understood that these provide:

• Powerful visualisation mediums
• Immersive environments
• Potential for highly interactive experiences
• Just look fantastic!
  

However such features do come with a few costs, including:

• A requirement for special software / skills = ($ & time)
• Issues with keeping complex models updated
• Hidden information in that (as shown above) surfaces can obscure each other
• And are generally not amenable to automation.
  

This is not to say that 3D is not a valid approach, in fact Spatial Vision has an award winning (Asia Pacific Spatial Excellence Award for Spatially Enabling Government) example of the use of 3D environments with geospatial information in a Groundwater Management visualisation tool developed for the Victorian Department of  Sustainability and Environment.

Profile Databases Concept

In developing the Profile Database concept we reviewed the design elements of the “classic” groundwater management tool – the Hydrogeological map. In general terms they are made up of:

• A planar map frame
• A profile (sometime two or three of these)
• Insert planar maps
• Graphs of thematic information
• Map marginalia

Profiles fit our design criteria to a tee, they:

• communicated the 3D or profile (“z”) relationships of the stacked surfaces
• are readily understood by the least map-centric users and are easily accessible to resource managers and the community.
  

Standard GIS environments have profiling functionality; however these tools are generally not designed with automation and mass production in mind.  Geoprocessing, automation via scripts (python), is the mechanism that enables development of large information-rich databases of profiles efficiently.

Profile Databases Terminology

Key terms associated with the development of a Profile Database include:

• Trace - the trace is the backbone of a particular set of profiles in a database, traces are defined (most often) by a line, however a point is also a valid Trace. 
• Transect - transects a lines that occur at intervals along the Trace or at angles around the Trace (point).
• Sample - these are locations along the Transect were the data stack is sampled and form the key “input” to the profiling process.

This hierarchy of concepts is welded together via a set of Unique Feature Identifiers that provide the basis for a logical and organised database.

Profile Databases Process

Fundamental to the development of a Profile Database from one or more surfaces is the conversion of planar geography (a graph of “x” and “y” coordinates) into profile geography (a graph of “d” and “z” coordinates, where:

X – standard planar coordinates in the east-west direction
Y – standard planar coordinates in the north-south direction
Z – coordinates in the profile (vertical) direction
D – distance along a particular transect

Each transect and the associated planar surfaces (in our case – aquifers) is processed using geoprocessing scripts (python) with the relevant “Z” and “D” information extracted, along with the relevant UFI information.  This “re-mapped” into the profile geography.  This information is progressively accumulated into the Profile Database
 

Having transformed the full set of profile into a database means that this derived dataset can now be used, analysed and visualised like any other GIS dataset.

The advantage of such a novel shift is to: 

• Reposition the profiles as a significant data resources as apposed to an output.
• Enable the profile to be interrogated and queried like any other GIS database
• Enable the profile to be attributed with contextual information like land use or tenure
• Provide a means to visualise the complex relationships that exist between aquifers
• Enable exaggeration in both the X and Y axis – most graphing approach only allow for Y exaggeration.  This one issue is significant where the vertical and horizontal scales of the data can have orders of magnitude difference.  The ability to manipulate both horizontal and vertical exaggeration enables vertical relationship to be emphasised without any detrimental impact on the generally understood shape of the surface in question.
• Enable the management of multiple transects or profiles in the one database environment.
• Enable the potential of linking Profile and Planar viewpoints for an interactive visualisation that is akin to a “standard” hydrogeological map product.

Desktop and web-based GIS technologies can provide the platform to link the planar data (transects) to the corresponding profiles providing a enhanced view and analytical capability for both groundwater professionals and the general public.

This concept has applications in a range of areas including operational management of irrigation channels by rural water authorities where it can be applied to mass generate cross-sectional profiles of channels for use in channel management packages such as HEC-RAS.  Spatial Vision has been investigating the development profile databases from LIDAR and high-resolution photography information with SRW.
 

Another application of Profile Databases realtes to coastal and estuary morphology. Spatial Vision is at an early in developing this arena however the results to date suggest a potential as a valuable tool to communicate the impacts of sea-level rise.
 

The concept of a profile database provides a powerful additional tool for natural resource managers to query, analyse, visualise and communicate the complex and inherently 3D or stacked information to internal and external stakeholders.

For further information contact: Milos Pelikan