Compilation of data related to Australian coal-mine fugitive emissions

This dataset quantifies characteristics of 62 Australian coal mines, with particular focus on quantities associated with fugitive emissions during the 2020-2021 Australian financial year.\n\nThe dataset is a compilation of data from 305 publicly-available data sources, which are mostly mine annual reviews, mine environmental impact assessments and government sources. Mining companies use methods specified in the Australian Government's National Greenhouse Accounts and National Inventory Reports to estimate their emissions. The 62 coal mines in the database are estimated to collectively produce 90% of fugitive emissions from currently-operating Australian coal mines.\n\nThe data are contained in the single SQL file combined_data.db. That database contains a data table for each mine with columns defining: (1) the parameters; (2) their values; (3) units; (4) notes related to the quantities; and (5) reference information. The database also contains a references table with information concerning the 305 primary data sources.\n\nTwo types of data are contained in combined_data.db: data directly sourced from the 305 publicly-available data sources; and quantities inferred from that data, which may be useful for fugitive-emissions modelling. The data sourced directly from the data sources are always provided with reference information. The inferred data are always provided with a brief description in the “note” column, and more details and calculations are found in the script analyse_data.py.\n\nCSIRO makes no claims concerning the accuracy of the data. Consumers of the data must decide whether the data are fit for their purpose. For the data directly sourced from the publicly-available data sources, the "note" column often contains remarks concerning supporting data (along with references) and sometimes contains remarks concerning apparently conflicting data (along with references) and the consumer of the data is encouraged to read the primary sources and consider the relevant notes. For the inferred data, consumers are encouraged to carefully consider whether the algorithms and models used are fit for their purpose.\n\nThe python script analyse_data.py may be used to produce various summaries (such as are presented in the data_description.docx file) as well as to combine the individual databases nsw_oc.db, nsw_ug.db, qld_oc.db and qld_ug.db (relating to opencut (OC) and underground (UG) mines in the Australian states of New South Wales (NSW) and Queensland (QLD)) into the single file combined_data.db. Most consumers of the data will not need to access these individual databases.\n\nThe data compares well with aggregated emissions in Australia’s National Greenhouse Accounts and National Inventory Reports, as well as the Safeguard Data published by the Australia's Clean Energy Regulator. The uncertainties in emissions are estimated in the National Inventory Reports to be ±10% for underground coal mines, and ±33% for opencut coalmines. Most other quantities in the database are also subject to uncertainty. For instance, the mining depth varies from year-to-year, from pit-to-pit, and longwall-panel-to-longwall-panel, but such detailed information is not usually provided in public data sources, and “mine depth” is a single number for each mine in the database. Similarly, coal-seam permeability is highly heterogeneous.\n\nLineage: Quantities from 305 publicly-available data sources, such as mine annual reviews, mine environmental impact assessments and government sources, were extracted to form the core of this dataset. Reference information is provided for all quantities.\nSecondly, the python script, analyse_data.py, uses those core numbers to infer various other, potentially useful, numbers. Each of these is provided with a note explaining their provenance, and the script analyse_data.py may be inspected to determine the algorithms used.

本数据集量化了62个澳大利亚煤矿的特征，尤其聚焦于2020-2021澳大利亚财年期间与逃逸性排放（fugitive emissions）相关的各项量化指标。

该数据集整合了来自305个公开数据源的信息，主要包括煤矿年度报告、煤矿环境影响评估及政府来源。煤矿企业采用澳大利亚政府《国家温室气体账户》（National Greenhouse Accounts）和《国家清单报告》（National Inventory Reports）中规定的方法估算其排放量。数据库中的62个煤矿据估计合计产生了澳大利亚当前运营煤矿逃逸性排放总量的90%。

数据存储于单一SQL文件combined_data.db中。该数据库为每个煤矿包含一个数据表，其列定义如下：（1）参数；（2）参数值；（3）单位；（4）与量化指标相关的注释；（5）参考信息。数据库还包含一个参考文献表，记录了305个原始数据源的相关信息。

combined_data.db包含两类数据：一类是直接来源于305个公开数据源的数据；另一类是从上述数据中推断出的量化指标，这些指标可能对逃逸性排放建模有用。直接来源的数据始终附有参考信息；推断数据则在‘note’列中附有简要说明，更多细节及计算过程可在脚本analyse_data.py中找到。

CSIRO对数据的准确性不作任何声明。数据使用者需自行判断数据是否符合其使用目的。对于直接来源于公开数据源的数据，‘note’列常包含关于支持性数据的说明（及参考文献），有时也包含关于明显冲突数据的说明（及参考文献），因此建议数据使用者阅读原始数据源并参考相关注释。对于推断数据，建议使用者仔细评估所用算法和模型是否符合其使用目的。

Python脚本analyse_data.py可用于生成各类汇总报告（如data_description.docx文件中所示），也可将多个独立数据库（包括新南威尔士州（New South Wales, NSW）和昆士兰州（Queensland, QLD）的露天煤矿（opencut, OC）及地下煤矿（underground, UG）数据库nsw_oc.db、nsw_ug.db、qld_oc.db和qld_ug.db）合并为单一文件combined_data.db。大多数数据使用者无需访问这些独立数据库。

本数据集与澳大利亚《国家温室气体账户》《国家清单报告》中的汇总排放量，以及澳大利亚清洁能源监管局（Clean Energy Regulator）发布的保障数据（Safeguard Data）具有良好的一致性。《国家清单报告》中估算的排放不确定性为：地下煤矿±10%，露天煤矿±33%。数据库中的大多数其他量化指标也存在不确定性。例如，开采深度随年份、矿坑及长壁工作面的不同而变化，但此类详细信息通常未在公开数据源中提供，因此数据库中每个煤矿的‘开采深度’仅为单一数值。类似地，煤层渗透率具有高度异质性。

数据溯源：从305个公开数据源（如煤矿年度报告、煤矿环境影响评估及政府来源）中提取量化指标，构成本数据集的核心。所有量化指标均附有参考信息。
其次，Python脚本analyse_data.py利用这些核心数据推断出其他各类潜在有用的量化指标。每个推断指标均附有说明其来源的注释，使用者可通过查看脚本analyse_data.py了解所用算法。