R code for Testing and implementation of the water quality metric for the 2017 and 2018 reef report cards (NESP TWQ 3.2.5, AIMS)

The dataset represents the code base developed for the generation of water quality metrics from various data sources (eReefs Biogeochemical models, MODIS Satellite imaging and AIMS in situ sampling).The water quality metric used underpinning previous Report Cards (until 2015) presented a number of significant shortcomings: - It was solely based on remote sensing-derived data. Concerns were raised about the appropriateness of exclusively relying on remote sensing to evaluate inshore water quality, considering well-documented challenges in obtaining accurate estimates from optically complex waters and the fact that only limited valid satellite observations are available in the wet season due to cloud cover; - It was limited to reporting on two indicators and did not incorporate other water quality data and indicators collected through the Marine Monitoring Program (MMP) and the Integrated Marine Observing System (IMOS); - It appeared relatively insensitive to large terrestrial inputs into the GBR lagoon during large rainfall and runoff events, most likely due to the binary assessment of compliance relative to the water quality guidelines and aggregation and averaging over large spatial and temporal scales;In 2016, based on the limitations described above, the Reef Plan Independent Science Panel (ISP) expressed a lack of confidence in the water quality metric that underpinned Report Cards (prior to 2015) and recommended that a new approach be identified for the Report Card 2016 and future Report Cards. The ISP also acknowledged substantial advancements in modelling water quality through the eReefs biogeochemical models and the fact that recent research and method development had improved our ability to construct report card metrics.Methods:- Index scoring strategies were systematically assessed to meet key objectives of sensitivity and representativeness, to allow data aggregation and to enable the integration of additional water quality measures when these become available in the future. A preferred method was identified as the scaled modified amplitude method with fixed caps sets at half and twice the threshold values, which was tested both theoretically and using historical datasets. - Aggregation strategies were reviewed and a hierarchical aggregation scheme was developed to allow multiple measures and sub-indicators to be combined into a single metric and to allow spatial and temporal aggregation. The process was designed to maintain the richness of information and allow the propagation of uncertainty, which were key project objectives.- The resulting water quality metric calculation process and parameters were applied to the development of the marine water quality metric component of Reef Report Card 2016, covering the reporting period 1 October 2015 to 30 September 2016. Format:The data are in the form of R code. Note, the code cannot be run as a standalone entity as it relies on non-public input data sources.本数据集为基于多源数据（eReefs生物地球化学模型、MODIS（Moderate Resolution Imaging Spectroradiometer）卫星遥感影像以及澳大利亚海洋科学研究所（AIMS）原位采样数据）生成水质指标所开发的代码库。 此前用于2015年及之前《珊瑚礁报告卡》的水质指标存在多项显著缺陷： - 仅依赖遥感衍生数据。考虑到光学复杂水体难以获取精准估算值，且湿季因云量覆盖导致有效卫星观测数据极为有限，仅依靠遥感评估近岸水质的合理性受到广泛质疑； - 仅涵盖两项水质指标，未纳入海洋监测计划（Marine Monitoring Program, MMP）与综合海洋观测系统（Integrated Marine Observing System, IMOS）所采集的其他水质数据与指标； - 在强降雨与径流事件期间，对大堡礁（Great Barrier Reef, GBR）潟湖的陆地大量输入物响应相对迟钝，这大概率源于其基于水质指南开展二元合规性评估，且在较大空间与时间尺度上进行了数据聚合与平均处理。 2016年，基于上述局限性，珊瑚礁计划独立科学专家组（Reef Plan Independent Science Panel, ISP）对2015年及之前支撑《珊瑚礁报告卡》的水质指标表示缺乏信心，并建议为2016年《珊瑚礁报告卡》及后续报告卡制定全新的指标构建方法。专家组同时认可了eReefs生物地球化学模型在水质建模领域取得的重大进展，以及近期研究与方法开发提升了我们构建报告卡指标的能力。 研究方法： - 系统评估了指数评分策略，以满足敏感性与代表性两大核心目标，实现数据聚合，并为未来新增水质指标的集成预留空间。最终优选出带固定阈值上下限（设定为阈值的一半与两倍）的缩放修正幅度法，并通过理论推导与历史数据集对该方法进行了验证。 - 对数据聚合策略进行了复盘，开发出层级聚合方案，可将多项指标与子指标整合为单一指标，同时支持空间与时间尺度上的聚合。该流程旨在保留信息丰富度并实现不确定性传递，这也是本项目的核心目标之一。 - 最终形成的水质指标计算流程与参数被应用于2016年《珊瑚礁报告卡》的海洋水质指标模块开发，报告覆盖周期为2015年10月1日至2016年9月30日。 数据格式： 本数据集采用R语言代码形式呈现。需注意，由于依赖非公开输入数据源，该代码无法独立运行。