The influence of schlieren on in situ optical measurements used for particle characterization
[Sequoia, July 7, 2008]
Mikkelsen OA, Milligan TG, Hill PS, Chant RJ, Jago CF, Jones SE, Krivtsov V, Mitchelson-Jacob G (2008): The influence of schlieren on in situ optical measurements used for particle characterization. Limnology & Oceanography:Methods 6:133–143.
Density differences, occurring for example in pycnoclines, can cause light scattering because the refractive index varies with density. This scattering – known as Schlieren – can occur even if no particles are present in the water.
Consequently the use of LISSTs in regions with strong density gradients may lead to problems in interpreting the data – are the data a result of scattering from particles or from density interfaces?
This problem can be answered if an independent way of getting particle size exists. Digital camera techniques offer such a way. In this work, Mikkelsen et al. measured the particle size and concentration with a camera and a LISST-100 in a pycnocline. The results show that both instruments are influenced by the density differences from the pycnocline: Schlieren causes automated image analysis routines to detect the Schlieren pattern as particles, while it causes the laser from the LISST to scatter intensely at small angles, thus being detected as large particles upon applying the inversion routine.
The images could be re-analysed manually by omitting the parts of the images affected by Schlieren. This demonstrated that there was in fact no increase in the size or concentration of particles in the pycnocline when compared to the size and concentration of particles above the pycnocline.
Buoyancy frequencies above 0.025 s-1 may cause increases in beam attenuation due to scattering from the density gradients and complete extinction of the laser beam occurred at buoyancy frequencies above ~0.20 s-1. All instruments measuring light scattering can potentially be affected by density differences, and results obtained in waters where buoyancy frequencies exceeds 0.025 s-1 should be interpreted carefully.