Spectroscopy has been used for many years for analytical purposes and its applications are numerous. Though the use of spectroscopy for analysis of water and salt water samples is not a novel concept, most research on the subject has been limited to the concentration range that would typically be found in sea water. Because of the similarities between NaCl and KCl, they have often been assumed to have few if any differences in their effects on the absorption spectrum of water. Correlations between salinity and absorbance have been developed up to about 10% of saturation using water absorption overtones but these results have not been extended to higher concentrations and have not focused on differentiation of species. The research demonstrated in this work showed that modelling the concentration of NaCl_(aq) and KCl_(aq) in water could be extended up to the saturation point of both salts and the concentrations of individual species could be estimated both in single-salt solutions and dual-salt solutions.

Spectra for water from 15°C to 95°C were used to show the effect of temperature on the absorbance spectra in the wavelength range of 180 nm to 1800 nm. A model was developed from these data that was capable of estimating temperature for water samples based on absorbance. This model was capable of estimating the temperature of water between 15°C and 95°C to within +/- 1.4°C using spectral measurements.

Temperature variation for salt water samples would have also provided useful data, however evaporation at these high temperatures presented a problem as sample concentration would change with loss of water. A method was designed to allow for future testing of salt-water samples at high temperatures that mitigated evaporation while maintaining atmospheric pressure.

Spectral data for NaCl_(aq) and KCl_(aq) in solution were collected for single-salt samples as well as samples containing both salts. These data were used to develop three models for determining concentration of both NaCl_(aq) and KCl_(aq) in solution. The first model was capable of differentiating between single-salt samples and determining the concentration of the solution with an average error of 0.9%. The second and third models were able to determine concentrations of both NaCl_(aq) and KCl_(aq) in dual-salt solutions. These models were able to correctly determine NaCl_(aq) and KCl_(aq) concentrations to within 3.0% average error for the second model and 2.4% average error for the third model. These models were tested using single-salt solutions between 90 g/L and 300 g/L and dual-salt solutions that would be typical in potash processing (90-160 g/L KCl_(aq) and 170-270 g/L NaCl_(aq)).