| In order to construct reasonable climate models that can predict weather patterns and assess the human impact on climate, in situ measurements of the ocean and the atmosphere are necessary. In order to track fluid movement, three variables are needed: temperature, pressure, and density. The first two are relatively easily, accurately, and economically measured in the ocean. The latter, density, is fundamental to typifying and predicting the path that waters will take, yet it is the hardest to measure. For nearly 200 years the measurement of seawater's salinity—salt content—has been used to determine its density to reasonable accuracy. Historically, this was achieved by precipitating seawater's Cl− with AgNO 3(aq)—the Mohr-Knudsen method—and empirically converting the result to salinity. This requires sample capture, and the accuracy is limited to 40 ppm. Thus, the chemical method was replaced in the 1960s with an empirically calibrated scale correlating seawater conductivity to salinity, sensitive to 1 ppm at best. Yet, conductivity, like Mohr-Knudsen, is not sensitive to all the matter within seawater that contributes to the density. The refractive index (RI) has been shown to correlate to density 80% better than conductivity. To date, no refractometers have been able to rival the sensitivity of conductivity probes for the all-desirable in situ measurement. A novel spectroscopic technique, known as Surface Plasmon Resonance (SPR), can measure the RI of a substance and lends itself applicable to fibre optic probes. The research presented herein shows that a fibre optic SPR probe can measure the salinity of a seawater analogue to a sensitivity in the 100 ppm range. This value is as good as or better than similar techniques recently presented in the literature and is more economical. With an easily attainable, higher-resolution spectrometer, this sensitivity is expected to approach the 10 ppm scale. Given that the cost of this probe is less than one-hundredth that of comparable conductivity probes, its lower sensitivity would be a reasonable trade-off for potentially wider application. Once the probe's response is calibrated and a drift prevented or corrected, field deployment would collect the necessary information for better climate modelling. |
