Periodic cleaning is required to remove bacterial films from instrument windows and particles that may not get flushed out of the backscattering enclosure. We have found, for typical oligotrophic, open-ocean conditions, that weekly cleaning is sufficient while, in meso- and eutrophic conditions, more frequent cleaning is needed. If, following cleaning, a significant change (drop) in signal is observed, fouling has likely degraded the previous data, which should be flagged accordingly and, if possible, corrected (for example by removing a trend). It is still unclear, however, whether one is better to assume a linear trend or an exponential trend, given that fouling organisms typically grow exponentially.
For cleaning details (e.g. solvent and/or soap types to be used) of specific sensors refer to manufacturer protocols. We recommend using lens paper on all the surfaces interacting with detected light (e.g. windows, flow sleeves) to insure that their properties do not change it time due to scraping with harsher materials. We typically use DI water, rubbing alcohol and a mild detergent such as Triton-X to clean our instruments.