"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"

Publish Australian Hydrological Geospatial Fabric (Geofabric) Data Product Via GeoServer

Introduction

The Geofabric database provides a seamless surface hydrography layer for Australia at a nominal scale of 1:250,000. It consists of lines, points and polygons representing natural and man-made features such as water courses, lakes, dams and other water bodies. The natural water course layer consists of a linear network with a consistent topology of links and nodes that provide directional flow paths through the network for hydrological analysis.

This activity will publish one of the Geofabric data product via GeoServer through Web Feature Service (WFS) and providing output in GeoSciML (GeoScience Markup Language).

Workflow (Target Date : End of December 2011)

  • Step 1 : Identify the Geofabric data product for publishing
  • Step 2 : Obtain SWIM GML application schema
  • Step 3 : Obtain Identified Geofabric Product's Dataset from BoM
  • Step 4 : Create Instance Documents base on SWIM GML application schema and BoM's Dataset
  • Step 5 : Create GeoServer configuration files base on the the instance documents
  • Setp 6 : Deploy GeoServer configufation files to BoM GeoServer instance in test server
  • Step 7 : GeoServer deployment guide

Step 1 : Identify the Geofabric data product for publishing

  • The Geofabric will evolve in phases over 10 years. The second phase, or Version 2.0 (released on 15 November 2011) comprises the following products :
    • Geofabric Surface Cartography - Cartographic representation of surface hydrological features
    • Geofabric Surface Network - Network representation of hydrological features
    • Geofabric Surface Catchments - Catchment boundaries derived from the 9 second Digital Elevation Model
    • Geofabric Groundwater Cartography - Cartographic representation of groundwater hydrology features
    • Geofabric Hydroolgy Reporting Catchments - Contracted nodes, contracted catchments and node-link network
    • Geofabric Hydrology Reporting Regions - Reporting regions developed for the Bureauís Australian Water Resources Assessment
  • Paul Box from Sustainable Water Information Models (SWIM) Project had indicated that Hydrology Reporting (HR) Catchments product will be used as this product contains the features and relationships that the AWRA-R team is interested in. Rob Atkinson from SWIM Project will complete and provide the HR Catchments application schema (app-schema) for ASRDC team by 30 September 2011.

Step 2 : Obtain SWIM GML application schema

Step 3 : Obtain Identified Geofabric Product's Dataset from BoM

  • The original HR Catchments product dataset (Version 2.0) is in ESRI GeoDatabase format and is available for download from the BoM ftp web site (ftp://ftp.bom.gov.au/anon/home/geofabric/).
  • Katie Mills from BoM had indicated that PostGIS database will be used instead of GeoDatabase and had provided a copy of PostGIS db dump for ASRDC to work on.

Step 4 : Create Instance Documents base on SWIM GML application schema and BoM's Dataset

  • Base on the HR Catchments application schema from SWIM Project and PostGISdatabase, the following instance documents had been created :
  • Some issues has been encountered while creating the instance document and solutions has been provided as below :
    1. With the current dataset, it is difficult to achieve unique gml:id for ahgf_cn:AHGFContractedCatchment feature type, the options available for gml:id are as below :
      • hydroID - unique across all products but doesn't persist thru different version
      • ConNodeIDs - persist thru different version but not unique and require combination with FConNodeID as composite key. However, not all record provide FConNodeID, base on explaination from Darren Smith of BoM : "!FConNodeID is only populated if the catchment has a corresponding link in the node-link network i.e. has FROM & TO (contracted) nodes. If the catchment has only a TO Contracted Node (ConNodeID) then FConNodeID remains unpopulated. In short, baring NoFlowAreas, every catchment has a contracted node to which it drains, some also have one or more upstream catchments which drain into them (FConNodeID being the draining contracted node for each of these).". A temporary solutions has been proposed as below to achieve uniqueness of gml:id until a new persistent ID (something like ConCatID) is avaialble.
      • Solution :
        if FConNodeID is not null, gml:id = 'ahgf_cn.AHGFContractedCatchment.' + ConNodeID + '.' + FConNodeID
        if FConNodeID is null, gml:id = 'ahgf_cn.AHGFContractedCatchment.' + ConNodeID 
    2. The relationship from AHGFContractedCatchment to AHGFLink to AHGFNode and back to AHGFContractedCatchment from the dataset doesn't match the design in app-schema. After some discussion with Darren Smith and Rob Atkinson, issues has been resolved and the final app-schema model is attached in Step 2.

Step 5 : Create GeoServer configuration files base on the the instance documents

Step 6 : Deploy GeoServer configufation files to BoM GeoServer instance in test server

Step 7 : GeoServer deployment guide

Topic attachments
I Attachment Action Size Date Who Comment
HR_Catchments_v3.PNGPNG HR_Catchments_v3.PNG manage 41.6 K 03 Nov 2011 - 15:33 FlorenceTan  
HR_Catchments_v4.PNGPNG HR_Catchments_v4.PNG manage 40.8 K 25 Nov 2011 - 13:41 FlorenceTan HR Catchments app-schema model
Topic revision: r9 - 13 Aug 2014, BenCaradocDavies