Abstract
Accurate ocean temperature measurement is critical for climate studies and ocean circulation modeling. Thermistor-based sensors, valued for high sensitivity and stability, require precise calibration due to their nonlinear resistance-temperature relationship. We evaluated six calibration equations (Basic, Hoge-1–4, S&H) across 273.15–308.15 K, and compared performance at calibration and non-calibration points. Furthermore, a novel temperature-distribution-optimized calibration strategy was proposed to enhance efficiency. Results showed that higher-order equations (Hoge-2, Hoge-3, Hoge-4, S&H) achieved sub-mK residuals with 13 calibration points, while lower-order equations failed. Extrapolation errors remained <1 mK within 2 K. With minimal calibration points, Hoge-2, Hoge-4, and S&H maintained <0.1 mK residuals at calibration points but exhibited increased residuals (~0.5 – 1 mK) in low-temperature non-calibration regions. The proposed strategy paired with the Hoge-2 equation reduced calibration time/cost by 50% compared to traditional 5 K-interval methods, without compromising accuracy. This work advances high-precision ocean temperature sensing, offering insights for in-situ calibration and climate model refinement.
Funding Information
  • Southern Marine Science and Engineering Guangdong Laboratory (GML2021GD0808)

This publication has 53 references indexed in Scilit: