“…Experiments demonstrate that the process of ice identification and INP counting are sensitive to illumination, camera resolution and image contrast, and that the INP identification procedure used until 2012 introduced asystematic errors. Thus, FRIDGE data generated between 2008 and 2012 (data appear in the following publications: Bundke et al, 2008;Klein et al, 2010a, b;Bingemer et al, 2012;Niemand et al, 2012) have been identified to include erroneous measurements, resulting from the miscounting of liquid droplets as ice. Unfortunately, the full extent of the error is variable and depends upon the sampled aerosol, meaning that the entirety of data collected from the stated time period and reported in the cited publications must be re-evaluated.…”