Click here to view this collection in the new DAP user interface
Murray-Darling Basin stream gauge daily data from 1990 to 2011, NetCDF format
These four NetCDF databases constitute the bulk of the spatial and spatiotemporal environmental covariates used in a latent health factor index (LHFI) model for assessment and prediction of ecosystem health across the MDB. The data formatting and hierarchical statistical modelling were conducted under a CSIRO appropriation project funded by the Wat... moreer for a Healthy Country Flagship from July 2012 to June 2014. Each database was created by collating and aligning raw data downloaded from the respective state government websites (QLD, NSW, VIC, and SA). (ACT data were unavailable.) There are two primary components in each state-specific database: (1) a temporally static data matrix with axes "Site ID" and "Variable," and (2) a 3D data cube with axes "Site ID", "Variable," and "Date." Temporally static variables in (1) include geospatial metadata (all states), drainage area (VIC and SA only), and stream distance (SA only). Temporal variables in (2) include discharge, water temperature, etc. Missing data (empty cells) are highly abundant in the data cubes. The attached state-specific README.pdf files contain additional details on the contents of these databases, and any computer code that was used for semi-automation of raw data downloads. less
Stochastic Analysis and Modelling
(1) For NSW I created the NetCDF database by (a) downloading CSV raw data from the NSW Office of Water real-time data website (http://realtimedata.water.nsw.gov.au/water.stm) during February-April 2013, then (b) writing computer programs to preprocess such raw data into the current format. (2) The same was done for QLD, except through the Queensland Water Monitoring Data Portal (http://watermonitoring.derm.qld.gov.au/host.htm). (3) The same was also done for SA, except through the SA WaterConnect => Data Systems => Surface Water Data website (https://www.waterconnect.sa.gov.au/Systems/SWD/SitePages/Home.aspx) during April 2013 as well as May 2014. (4) For Victoria I created the NetCDF database by (a) manually downloading XLS raw data during November and December in 2013 from the Victoria DEPI Water Measurement Information System => Download Rivers and Streams sites website (http://data.water.vic.gov.au/monitoring.htm), then (b) writing computer programs to preprocess such raw data into CSV format (intermediate), then into the current final format.
Additional details on lineage are available from the attached README.pdf files.
The creation of these NetCDF databases was part of G Chiu's appropriation project from July 2012 to June 2014 funded by the CSIRO Water for a Healthy Country (flagship) -> Ecosystems and Contaminants (theme) -> Ecosystem Responses to Flow (stream). The lineage information above includes the sources of raw stream gauge data from which these databases were created. G Chiu thanks Dr C Pollino (ERF Stream Leader) and Ms L Merrin (CSIRO Land and Water) for their valuable help towards locating and understanding the raw data; Ms Anne Stevenson, Mrs Sue Cook, and Mr Dominic Hogan (CSIRO Research Data Support) for their help to coordinate the legal clearance and public access of these databases through the CSIRO Data Access Portal; and Ms Luisa Ermacora (Victoria Department of Environment and Primary Industries) for coordinating early legal clearance for the public release of my Victoria NetCDF database.
Creative Commons Attribution 3.0 Unported Licence
Chiu, Grace (2014): Murray-Darling Basin stream gauge daily data from 1990 to 2011, NetCDF format. v3. CSIRO. Data Collection.
All Rights (including copyright) CSIRO Australia 2014.
The metadata and files (if any) are available to the public.
TECO 1136.5 Spacio-temporal modelling
This was a methodological research project, aimed to develop a spatial statistical modelling approach that is tailored to assess and predict how ecosystem health across the MDB responds to changes in flow and other environmental attributes. A fully developed model would mitigate difficulties in existing practices given (i) spatially sparse biotic d... moreata and (ii) the need for rigorous uncertainty assessment. This new integrated modelling approach is based on the statistical LHFI methodology (see above Related Materials), and is tailored to exploit the accessibility and spatial nature of geomorphological and other attribute data to handle (i)-(ii) over space. Thus, compared to conventional, non-model-based health scoring methods, the spatial/spatio-temporal LHFI method could improve one's ability to interpolate health conditions and predict their response to perturbations of environmental attributes such as flow, without the need to revamp standard field data sampling protocols. less
Environmental covariates for modelling MDB ecosystem health
Many spatial and spatiotemporal environmental covariates are drivers of ecosystem health across the MDB. Stream gauges collect data on a long list of variables, many of which are candidate covariates for ecosystem health modelling. Each jurisdiction disseminates their stream gauge data to the public in different formats. Raw stream gauge data, alon... moreg with basic geospatial metadata, were downloaded from state government websites, then collated and aligned to form these four NetCDF databases. They constitute the bulk of the environmental covariates used in this project. less
See attached README.pdf files. (Note: Release v2 and thereafter of this data product differ from v1 only in the file type of README files, being changed from RTF to PDF.)