Adelaide Plains Groundwater Model 2015: Assessment of Adelaide Plains Groundwater Resources

Model produced as a result of GOYDER INSTITUTE FOR WATER RESEARCH Project No. I.1.6. This numerical model simulates groundwater flow in steady-state (long-term historical state under pre-development conditions – no pumping) and in different transient conditions: 1900-1950 under pre-development conditions (no pumping); 1950-2013 under development conditions (pumping); 2013-2100 under different predictive scenarios (increase/decrease of pumping, climate change, increase of MAR). The model also simulates chloride and carbon-14 transport, both in steady-state (long-term historical state under pre-development conditions – no pumping).<br><br>The data upon which the model depends are stored together with the model, which can therefore be run “as is”. The data can potentially be updated (using updated versions of the sources described above, for example) but the model would need to be updated accordingly. Updating the model with new data is relatively easy using the Matlab scripts as long as the data keep the same structure. Note that anything affecting the historical simulations implies that the model would have to be recalibrated.<br><br>Data Sources and Modelling Software<br>Modelling Software:<br>- This is a MODFLOW-based model. MODFLOW and related programs are freely available for download from http://water.usgs.gov/ogw/modflow/.<br>- Pre- and post-processing was done in Matlab (commercial product; see http://au.mathworks.com/) using the mfLab toolbox (freely available for download from https://code.google.com/p/mflab/).<br>- Some data pre-processing was also done in Python (freely available for download from https://www.python.org/) using the ArcPy package that comes with ArcGIS (commercial product; see http://www.esri.com/software/arcgis).<br>- Calibration and uncertainty analysis was done with PEST (freely available for download from http://www.pesthomepage.org/).<br>Data Sources:<br>- Surface water features, groundwater levels and chloride data were sourced from the WaterConnect database managed by DEWNR (https://www.waterconnect.sa.gov.au/). This data can be downloaded at no cost and is licensed under a Creative Commons Attribution 3.0 Australia Licence.<br>- Hydrogeologic units extent and elevation were sourced from DEWNR (personal communication, no licence; contact at DEWNR: Steve Barnett).<br>- 14C data was sourced from the literature and appended by new data acquired during the project (no licence).<br>- Historical pumping and managed aquifer recharge (MAR) data were sourced from the previous modelling effort of the same area by RPS Aquaterra (Georgiou J, Stadter M, Purczel C (2011) 'Adelaide Plains groundwater flow and solute transport model'). The previous model is owned by DEWNR (contact at DEWNR: Juliette Woods).<br>- Rainfall data was sourced from the SILO database (https://www.longpaddock.qld.gov.au/silo/) supplied under licence to Flinders University with a right to distribute digital products based on or containing the data, as this model.<br>- Rainfall projection data sourced from Enviro Data SA (https://data.environment.sa.gov.au/Climate/SA-Climate-Ready/). This data can be downloaded at no cost and is licensed under a Creative Commons Attribution 4.0 International Licence.<br>- Future MAR scheme data were provided by DEWNR (personal communication, no licence; contact at DEWNR: Graham Green).<br><br><br>Location: The Adelaide Plains, extending from the major faults at the foothill of the Mount Lofty Ranges in the south and east (Eden-Burnside and Para faults), up to 5 km offshore in the west, and bounded by the Light River in the north. In the vertical direction, the model extends from the land surface down to the bedrock included (model bottom fixed at -700 mAHD).

本模型为戈耶德水研究所（GOYDER INSTITUTE FOR WATER RESEARCH）I.1.6号项目的产出成果。该数值模型可模拟稳态（开发前无抽水的长期历史状态）与不同瞬态条件下的地下水流运动：1900-1950年的开发前无抽水状态、1950-2013年的开发抽水状态，以及2013-2100年的多情景预测状态（抽水规模增减、气候变化、含水层人工补给（Managed Aquifer Recharge, MAR）变化）。此外，模型还可模拟稳态（开发前无抽水的长期历史状态）下的氯离子与碳-14运移过程。

模型依托的数据集与模型本体一同存储，可直接“开箱即用”。数据可基于更新后的来源（例如上述提及的数据源的新版本）进行更新，但需同步对模型进行适配调整。只要数据结构保持一致，通过Matlab脚本即可较为便捷地完成模型的数据更新。需注意：任何影响历史模拟的修改，均需对模型重新进行校准。

数据来源与建模软件
建模软件：
- 本模型基于MODFLOW开发。MODFLOW及其相关程序可从http://water.usgs.gov/ogw/modflow/免费下载获取。
- 模型的前后处理工作依托商业软件Matlab，可通过http://au.mathworks.com/了解详情，并使用mfLab工具箱完成，该工具箱可从https://code.google.com/p/mflab/免费下载。
- 部分数据预处理工作通过Python完成，可从https://www.python.org/免费下载，搭配ArcGIS自带的ArcPy包实现，ArcGIS为商业软件，详情可访问http://www.esri.com/software/arcgis。
- 模型校准与不确定性分析借助PEST完成，该工具可从http://www.pesthomepage.org免费下载。

数据来源：
- 地表水要素、地下水位与氯离子数据均取自DEWNR管理的WaterConnect数据库（https://www.waterconnect.sa.gov.au/）。该数据可免费下载，采用知识共享署名3.0澳大利亚许可协议（Creative Commons Attribution 3.0 Australia Licence）授权。
- 水文地质单元的分布范围与高程数据取自DEWNR（私人沟通获取，无特定许可协议；DEWNR联系人：Steve Barnett）。
- 碳-14数据取自公开文献，并补充了本项目期间获取的新数据（无特定许可协议）。
- 历史抽水与含水层人工补给（MAR）数据取自RPS Aquaterra对同一区域此前的建模成果（Georgiou J, Stadter M, Purczel C (2011) 《阿德莱德平原地下水流与溶质运移模型》）。该既往模型归DEWNR所有（DEWNR联系人：Juliette Woods）。
- 降雨数据取自SILO数据库（https://www.longpaddock.qld.gov.au/silo/），该数据已获得授权许可给弗林德斯大学（Flinders University），允许基于该数据分发衍生数字产品，本模型即属于此类产品。
- 降雨预测数据取自Enviro Data SA（https://data.environment.sa.gov.au/Climate/SA-Climate-Ready/）。该数据可免费下载，采用知识共享署名4.0国际许可协议（Creative Commons Attribution 4.0 International Licence）授权。
- 未来含水层人工补给（MAR）方案数据由DEWNR提供（私人沟通获取，无特定许可协议；DEWNR联系人：Graham Green）。

研究区域为阿德莱德平原：南、东以洛夫蒂岭山麓的主要断层（伊甸-本戴德断层与帕拉断层）为界，西向延伸至离岸5公里处，北以莱特河为边界。垂向上，模型覆盖范围从地表延伸至基岩，模型底部固定在海拔-700米澳大利亚高程基准（mAHD）处。