Thermal-optical analysis (TOA) using a transmittance protocol is the reference method for the determination of elemental and organic carbon collected on ambient air filters (DIN EN 16909:2017). In Europe, EUSAAR2 is the standard protocol used for these samples (Cavalli et al, 2010). The method is also commonly applied to quantify water insoluble organic and elemental carbon in snow samples after filtration. Fe-oxides in mineral dust deposited on the sample filter can interfere with the quantification of elemental carbon, leading to a systematic bias or even erroneously simulating that elemental carbon is not present in the sample at all. For remote environments like glaciers, the determination of elemental carbon and mineral dust is of special interest, as they reduce the albedo of the snow surface after deposition leading to various adverse effects, e.g. premature melting of the snow cover. A first indication of interference by mineral dust is an increase in laser transmittance during the calibration phase, when the temperature drops, as the optical properties of these Fe-oxides are temperature dependant. Figure 1, showing duplicate analysis of one filter sample, illustrates this behaviour, which was described earlier (Kau et al, 2022). Comparing the two analytical runs, the change in transmitted laser signal is much larger for the first run, when pyrolysis and combustion of carbonaceous material occurs. Figure 1. Transmitted laser signal of two consecutive analytical runs of a PM₁₀ sample collected at Sonnblick Observatory during a mineral dust event. We discuss which consequences and possibilities arise: (1) the necessary correction of the transmitted laser signal to obtain a correct split point for the determination of organic and elemental carbon and (2) an approximation of the mineral dust loading using the Fe loading derived from TOA data. In the years 2021 and 2022, 13 % of PM₁₀ filters collected weekly at the Global Atmosphere Watch station Sonnblick Observatory showed signs of possible interference. For snowpacks sampled at the same site, representing deposition of the previous accumulation period, between 10 and 38 % of samples within the snowpacks showed these noticeable features. We propose a correction, which uses transmittance data of two consecutive analytical runs. The first run is influenced by carbonaceous compounds and mineral dust. While carbon is completely removed during the first run, refractory mineral dust is left on the filter and influences the laser transmittance of the second run as well. We apply the correction to raw data of PM₁₀ and filtrated snow samples to minimize the bias in the quantification of elemental and organic carbon introduced by mineral dust. We discuss the impact of the correction on the results obtained for organic and elemental carbon. An approximation of the mineral dust loading from the Fe loading necessitates a factor describing the mass fraction of Fe and mineral dust arriving at Mount Sonnblick after long-range transport. We deduce this factor from elemental data of sample filters and assess the mineral dust loading directly from TOA data. We further compare our results with approaches described in literature using several elements.