Hi, I am not quite sure about the working principle by which the presence of water is detected. I believe that is due to water absorption @ a specific frequency of the water (say, the one water absorbs the most). However, this would mean that the Melexis IR thermometer you are using should be able to measure a specific frequency of the spectrum. Did you use pre-filtering? Or was it post-processing? I guess that in your the light used as "input" to the water in your case was the natural one, and one could do the same with a LED, having a more focused output information. Thanks a lot!