Ice crystal formation in atmospheric clouds has a strong effecton precipitation, cloud lifetime, cloud radiative properties, and thus theglobal energy budget. Primary ice formation above 235 K is initiatedby nucleation on seed aerosol particles called ice-nucleating particles(INPs). Instruments that measure the ice-nucleating potential of aerosolparticles in the atmosphere need to be able to accurately quantify ambientINP concentrations. In the last decade several instruments have beendeveloped to investigate the ice-nucleating properties of aerosol particlesand to measure ambient INP concentrations. Therefore, there is a need forintercomparisons to ensure instrument differences are not interpreted asscientific findings.In this study, we intercompare the results from parallel measurements usingfour online ice nucleation chambers. Seven different aerosol types are testedincluding untreated and acid-treated mineral dusts (microcline, which is aK-feldspar, and kaolinite), as well as birch pollen washing waters.Experiments exploring heterogeneous ice nucleation above and below watersaturation are performed to cover the whole range of atmospherically relevantthermodynamic conditions that can be investigated with the intercomparedchambers. The Leipzig Aerosol Cloud Interaction Simulator (LACIS) and thePortable Immersion Mode Cooling chAmber coupled to the Portable IceNucleation Chamber (PIMCA-PINC) performed measurements in the immersionfreezing mode. Additionally, two continuous-flow diffusion chambers (CFDCs)PINC and the Spectrometer for Ice Nuclei (SPIN) are used to performmeasurements below and just above water saturation, nominally presentingdeposition nucleation and condensation freezing.The results of LACIS and PIMCA-PINC agree well over the whole range ofmeasured frozen fractions (FFs) and temperature. In general PINC and SPINcompare well and the observed differences are explained by the ice crystalgrowth and different residence times in the chamber. To study the mechanismsresponsible for the ice nucleation in the four instruments, the FF (fromLACIS and PIMCA-PINC) and the activatedfraction, AF (from PINC and SPIN), are compared. Measured FFs are on theorder of a factor of 3 higher than AFs, but are not consistent for allaerosol types and temperatures investigated. It is shown that measurementsfrom CFDCs cannot be assumed to produce the same results as those instrumentsexclusively measuring immersion freezing. Instead, the need to apply ascaling factor to CFDCs operating above water saturation has to be consideredto allow comparison with immersion freezing devices. Our results providefurther awareness of factors such as the importance of dispersion methods andthe quality of particle size selection for intercomparing online INPcounters.
展开▼
