The extraction process finds the model grid cell that is closest to the specified extraction coordinate. This grid cell is then used to extract the data. For the the 4 km model this means that the model data may be up to close to 2.8 km (if it is near the corner of a grid cell) from the specified location.

The eReefs models are numeric simulations that occur on a 3D grid. They simulate changes over time based on the physics programmed into the model and the boundary conditions that drive changes. The outer boundary of the 4 km hydrodynamic model is driven by ocean currents estimated from lower resolution global models. The 1 km hydrodynamic model outer boundary is driven from the 4 km hydrodynamic model. This nesting of finer and finer models makes sure that there is not a large jump in resolution at the boundary. The ratio of model should be no more than 4:1 to ensure the boundary does not introduce anomalies into the model.

This nesting is why there is both 4 km and 1 km resolution models.

The 1 km model requires a lot more computations than the 4 km models. It has more than 10 times more pixels in the model grid and due to its finer grid requires a corresponding increase in number of time steps that need simulating. When the eReefs project started (2015) the 1 km model took almost 1 day to run 1 day of simulation, and thus if the start date was set in the past then the model would never catch up. Since the 4 km model ran much faster its start day could be further in the past (2010).

The eReefs models are nested so that the 1 km resolution model is driven from the 4 km resolution model. If there are errors in the 4 km model then they will be carried into the 1 km model and thus it is important that the 4 km model is as accurate as possible. For this reason there has been much more research focus on tuning the 4 km model than the 1 km version. The 4 km model takes vastly less time to compute and so all recent improvements have been made to this version of the model.

The eReefs models are the result of over 25 years of research. They represent the most detailed and studied hydrodynamic and biogeochemical models of the Great Barrier Reef region. Details of these models along with an assessment of the biogeochemical model against in-situ measurements can be found on the CSIRO eReefs Research Models site. An overview and skill assessment is provided in the eReefs: An operational information system for managing the Great Barrier Reef.

In general the hydrodynamic model is more accurate than the biogeochemical model as it is a physics based simulation of the water movement. The biogeochemical model is driven by the hydrodynamic model and incorporates any errors in the hydrodynamics. Biogeochemical model variables such as chlorophyll are particularly difficult to model accurately as they are the result of a long change of processes (hydrodynamics, sediment, resuspension, nutrient recycling, etc.) that each must be accurate to get an accurate final simulation. As a result eReefs model outputs should be used as an additional source of information and should not be used without consideration of where it is fit for the analysis being undertaken.

The GBR4 BGC version 3.1 was setup for scenario comparisons. River loads are based off 2019 land practices over the complete history (2010 - 2019). Using this version to estimate historic water quality parameters needs to be done with caution.

The data is extracted from the original model data and aggregated to the time period specified by the extraction. The eReefs model data is very large (15 TB+) resulting in considerable moving around of data to extract the relevant data. The extraction process can take a couple of hours. As a result the extraction will not be immediately available. You will be notified by email when the extraction is complete.

Each data extraction creates a status page showing the progress of the extraction and the resulting data download. This page can be provided as a reference to others to access the data. The reference for the data should cite

CSIRO ({Year}) eReefs {Name and version of the model}. [data file]. Retrieved from {Extraction page URL}

For example:

CSIRO (2020) eReefs GBR 4 hydrodynamic model version 2. 2001.1298f22b-summary.csv. Retrieved from https://extraction.ereefs.aims.gov.au/data/2001.1298f22b

This data extraction tool was developed by the Australian Institute of Marine Science. It operates on the eReefs model data developed and maintained by CSIRO.

The intention is that the extraction status pages are a permanent link to the data for download. As a result you can safely use these links in documents. If there are any URL changes in the future then these links will be maintained with URL redirects.

This tool is designed to allow data extraction at specific locations. It has a limit of 2000 locations per extraction. If you need to perform grid level data processing it is probably best to download and process the original raw NetCDF data.

The data extracted from this service is made available under a Creative Commons Attribution license. If the data is used in the paper then it should be cited following the guidance above in 'If I use this data in a paper how should I cite the data?'

Each data extraction creates a status page with the data available for download. Each of these pages is publicly available, but only to those that know the random URL.