"Seegrid will be due for a migration to confluence on the 1st of August. Any update on or after the 1st of August will NOT be migrated"

A model for Landslide data

Background on landslide inventories

Currently a number of different landslide inventories exist in various databases within Australia, and each uniquely addresses a specific purpose. These databases range in scale and detail, and although some similarities and a number of common themes are apparent between databases, the method in which information is organised and described varies considerably. This means information cannot readily be compared or aggregated with other sources. Furthermore, these inventories are generally only accessible to a small number of individuals and subsequently, it is possible there is significant duplication of effort among landslide researchers independently attempting to fill information gaps.

Landslide inventories are fundamental for developing rigorous hazard and risk assessments. However, an agreed, systematic way of developing these inventories is presently not available, neither is there an example of ‘best practise’ that could be used as a guideline in Australia.

Landslide Database Interoperability Project (LDIP)

The LDIP is a pilot project that was established as part of a continued endeavour between Geoscience Australia (GA), Mineral Resources Tasmania (MRT) and the University of Wollongong (UoW) to improve the historic record of landslide events in Australia and to ensure information is useful and relevant, as well as widely available and accessible.

This project demonstrates how three diverse landslide inventory databases being maintained independently by different organisations at different scales, can be seamlessly accessed via a standardised web interface as one ‘virtual’ database. The project highlights the functionality and benefits of adopting this approach to manage and access landslide information in real time as well as the advantages in developing an agreed framework for the database structure of landslide inventories which adopts existing standards and recommended classification systems,

Therefore, the primary goals of the Landslide Database Interoperability Project are:

  1. To demonstrate interoperability between landslide inventories managed by a number of organisations
  2. To develop a framework for landslide inventories that can be considered as ‘best practise’ in Australia.

It is expected that interoperability between landslide databases in Australia will be demonstrated by the end of the current financial year. However, the development of an agreed framework for landslide inventories will continue into the coming financial year. Database custodians, academia, landslide experts, interoperability experts, representatives of various AGS working groups, as well as some overseas agencies are contributing to the development of the inventory framework. Consultations are still underway at present and it is expected they will continue into 2008.

Therefore, the demonstrator project will use a preliminary model of this framework. It is anticipated the landslide data model will be refined further in the coming financial year when further progress is made in establishing an agreed, collaborative framework for landslide inventory databases (subject to funding).

The virtual database will be subject to review and comment by users once the demonstrator is online. This will assist us in prioritising database refinement, the development of database extensions and highlight the areas where additional functionality is required.

Benefits of adopting an interoperable approach for landslide inventories Interoperability will enable landslide information to be accessed in real time by all levels of government, geotechnical professionals, emergency managers, land use planners, academics and the general public regardless of where it is hosted. It provides direct access to spatial-enabled data and allows users to simultaneously search and query the most up-to-date information available in geographically distributed databases through a single website. The search results can be displayed as reports, graphs, maps, statistics or tables, and data can be queried against background datasets, such as topography, geology and geomorphology.

The virtual database has the capability to display and manage site specific details as well as broader-scale information. There will be no limit to the number of landslide databases that can be accessed via the interface, and each individual database will be maintained in its current format by its original custodian. These database custodians will also have the ability to restrict access to particular database attributes that contain information considered sensitive or confidential. The functionality of interoperability allows the virtual landslide database to be shaped and tailored to specific user needs.

Meeting the needs of decision makers

It was important to understand who the different users of landslide data are, and what level of information they require for decision making. It was also important to determine whether the type of data currently being collated in inventories is useful to decision makers, and whether the value of these inventories could be increased with some additional data capture, where available.

A research and analysis phase was conducted as part of the LDIP to address some of these questions. It was found that information presently captured meets the requirements of some users, but not all. Furthermore, each private database has a unique way of describing information, which raised important questions regarding the design of the interface, such as “is there a standard way to describe a landslide event?” and “should there be a standard way to describe an event?” The Australian Geomechanics Society (AGS) have pioneered many landslide hazard initiatives in Australia, and due to these initiatives we believe we are at a stage where agreement can be reached on how to describe landslide events within landslide inventories.

Subsequently, this data model will adopt, and encourage the use of agreed standards, classification systems and terminology to describe landslide events. This model will also provide the basis for accessing further landslide inventories via the interface in the future.

Current state of play:

Geoscience Australia has been coordinating the development of a collaborative landslide inventory framework as part of the Landslide Database Interoperability Project (LDIP). The inventory structure adopts agreed standards, terminology and classification systems as outlined within Landslide Risk Management Guidelines developed by the Australian Geomechanics Society (AGS) in 2000.

The framework aims to provide guidance in the fields required for reporting. Fields chosen have been selected with consideration of data requirements for landslide risk assessment and aim to highlight where current gaps in knowledge are. The framework is being developed in conjunction with project partners, members of the AGS Landslide Risk Management taskforce, review committees and local government. It is anticipated the inventory framework will be provided to the AGS for technical review in 06/07.

Once agreement has been reached on this inventory framework, the landslide data model described here, will be revised to adopt any changes to the framework and feedback from the community.

This demonstrator project is anticipated to be the first step of a bigger picture for the way in which natural hazard information (and other information) is organised within the agency. The landslide database will be an ongoing process of development and possible extensions of this pilot study are being considered for 07/08.

Future goals and vision

An interactive data entry tool may be developed at a later stage to further encourage the reporting of landslides. It is envisaged this tool, through a registration process, will enable users to submit and use quality-controlled landslide data via the website, without needing to establish and maintain their own databases. It is recognised that different types of investigations allow different levels of reporting, which will be taken into account in the development of this tool, and the data model.

It is envisaged that this landslide data model, once further refined, could become a national template for the ‘best practise’ capture of landslide information. At a national level, landslide data is currently captured on an ad hoc basis and this initiative will highlight data that is most valuable to decision makers, whilst providing the basis, and ability for individuals establishing new landslide inventories to easily link into the interface.

If landslide database developers adopt this proposed ‘best practise’ data model, and if the landslide community assist in populating these databases and contribute relevant information to database custodians, it is feasible that in the future and with appropriate planning, there could be one virtual landslide database, with national coverage of the best and most up to date information which is widely accessible and tailored to meet the needs of various users and decision makers. More importantly, it will ensure that information on landslide hazard and risk is accessible to those who need it.

Intended users

Intended users of the virtual database include: federal government agencies seeking information on the cost of landslides to people and property; state agencies seeking information on the location of landslides and their characteristics, in order to focus mitigation efforts and regional land use planning; local agencies may seek more detailed information for use by town planners, emergency managers and geotechnical practitioners for land use planning and hazard zonation.

As a pilot focused on interoperability issues, this project will not exhaustively engage with possible users in the demonstrator phase, but instead rely on expert input to identify the typical types of interactions that might be required.

Further information

This virtual database will be an ongoing process of development, and it is important to note that when the demonstrator goes online, the database will not contain all necessary and readily captured information required by users. The first step is to demonstrate the value in adopting an interoperable approach. The virtual database and landslide data model will be refined with time after this initial step. Nonetheless, the LDIP will provide the opportunity for the community to start capturing relevant information in a consistent way to help various users and decision makers involved in reducing landslide risk to Australian communities. It is hoped this virtual database may eventually become a one-stop resource for landslide event information in Australia.

This project aligns well with recommendations made by the Council of Australian Governments (COAG) in the COAG Report: Natural Disasters in Australia: reforming mitigation, relief and recovery arrangements. The report recognises that a more consistent approach to data research and analysis is needed across all levels of government for a greater understanding of natural disaster mitigation. Many of the 66 recommendations in the report advocate a comprehensive approach to managing natural disasters and, in particular, Reform Commitment 2; “Establish a nationally consistent system of data collection, research and analysis to ensure a sound knowledge base on natural disasters and disaster mitigation”. This report is available for download at: http://www.ema.gov.au/www/emaweb/emaweb.nsf/Page/Publications_ProgramPublications_CouncilofAustralianGovernments_COAGReportonNaturalDisastersinAustralia

Notes on the data model

The data model presented here is a preliminary model based on a draft iteration of the landslide inventory framework that was current in March 07.

The data model shown here will be updated once the demonstrator project is nearer to completion.

Accessing the data model

The data model is being developed by RobAtkinson, SimonCox, NickArdlie, StuartGirvan and LesleyWyborn with scientific rigour and guidance being coordinated by MonicaOsuchowski.

The model has been developed as a GML Application Schema and utilises aspects of the ISO harmonised model and OGC SWE packages (namely Observation & Measurements and Sampling). It is currently managed under Subversion control.

Read-only access is possible using the following steps:

For more information on the methodology associated with GML Application Schema development and the relationship between the source UML models and the XML implementation models see SchemaFormalization.

A preliminary set of auto-generated HTML documentation of the model is available below:

For further information on the LDIP please contact: MonicaOsuchowski

-- MonicaOsuchowski - 08 Mar 2007
Topic attachments
I Attachment Action Size Date Who Comment
landslides-html-doc.zipzip landslides-html-doc.zip manage 159.9 K 14 Mar 2007 - 10:13 NickArdlie Auto-generated HTML documentation of the model. (Preliminary version only)
Topic revision: r8 - 15 Oct 2010, UnknownUser
 

Current license: All material on this collaboration platform is licensed under a Creative Commons Attribution 3.0 Australia Licence (CC BY 3.0).