Statistical Calibration of a Portable pH Sensor for Coastal Monitoring: A Case Study in Mauritius

The application of sensors for measuring physico-chemical parameters like sea surface temperature and pH enables rapid, reliable long-term data collection. This is particularly crucial for Small-Island Developing States, whose marine ecosystems are vulnerable to climate change and ocean acidification. This study compared the pH data recorded by a portable pH sensor to those determined in the laboratory (UV-Vis spectrophotometric method) at Flic-en-Flac, a lagoon in the west coast of Mauritius. The two methods were compared using the Bland-Altman analysis and a multiple linear regression. It was noted that without adjusting for temperature differences between the pH sensor and a temperature probe, the pH results indicate a Pearson’s correlation coefficient of 0.473 (r2 = 0.224, p = 0.00135). In contrast, when temperature discrepancies between the pH sensor and the temperature probe were accounted for, the two methods yielded comparable results (Pearson’s r = 0.919, r2 = 0.959, p < 0.001). With the temperature adjustments, the data provided by the pH sensor are considered reliable and can be a complementary method to laboratory-based pH measurement.