Glider vehicles are now perhaps some of the most prolific providers of real-time and near-real-time operational oceanographic data. However, the data from these vehicles can and should be considered to have a long-term legacy value capable of playing a critical role in understanding and separating inter-annual, inter-decadal, and long-term global change. To achieve this, we have to go further than simply assuming the manufacturer’s calibrations, and field correct glider data in a more traditional way, for example, by careful comparison to water bottle calibrated lowered CTD datasets and/or “gold” standard recent climatologies. In this manuscript, we bring into the 21st century a historical technique that has been used manually by oceanographers for many years/decades for field correction/inter-calibration, thermal lag correction, and adjustment for biological fouling. The technique has now been made semi-automatic for machine processing of oceanographic glider data, although its future and indeed its origins have far wider scope. The subject of this manuscript is drawn from the original Description of Work (DoW) for a key task in the recently completed JERICO-NEXT (Joint European Research Infrastructure network for Coastal Observatories) EU-funded program, but goes on to consider future application and the suitability for integration with machine learning.

中文翻译：

---

使用 Theta/S 数据的空白最大化图像分析对滑翔机盐度数据进行近乎自动的常规现场校准/校正

滑翔机现在可能是实时和近实时操作海洋数据的最丰富的提供者之一。然而，来自这些车辆的数据可以而且应该被认为具有长期的遗产价值，能够在理解和区分年际、年代际和长期全球变化方面发挥关键作用。为了实现这一目标，我们必须走得更远，而不是简单地假设制造商的校准，并以更传统的方式现场校正滑翔机数据，例如，通过与水瓶校准的降低 CTD 数据集和/或“黄金”标准最近的气候进行仔细比较。在这份手稿中，我们将一项历史技术带入 21 世纪，该技术已被海洋学家手动使用多年/数十年，用于场校正/相互校准、热滞后校正、和生物污染的调整。该技术现在已经被半自动地用于海洋滑翔机数据的机器处理，尽管它的未来和它的起源有更广泛的范围。本手稿的主题来自于最近完成的 JERICO-NEXT（欧洲海岸观测站联合研究基础设施网络）欧盟资助计划中的一项关键任务的原始工作描述 (DoW)，但继续考虑未来的应用和与机器学习集成的适用性。