Short-term Variation in Measurements of Atmospheric Ice-Nucleating Particle Concentrations

Primary ice formation in clouds is initiated by a minor subset of the total atmospheric aerosol population called ice-nucleating particles (INPs), that originate from various source regions around the globe. Tropospheric clouds containing ice crystals are a key player for facilitating life on Earth by providing fresh water originating from liquid and solid precipitation and by balancing the radiation budget with their optical properties. The number concentration of atmospheric INPs and the related number of ice crystals is highly variable and strongly depends on the aerosol type and properties. Due to changes in the ambient conditions or physical or chemical processes in the atmosphere on a time scale of minutes to a few hours, the INP concentration may vary on a similar time scale. Especially the potential short-term variation is still not well investigated and underrepresented in models, but may be addressed with automated INP measurements with a high time resolution at well controlled conditions. A major subject of the here presented thesis is to quantify the short-term variation of atmospheric INP concentrations and search for possible causes for this variation by correlating the INP measurements with other variables like the aerosol number and surface area concentration and meteorological variables.
... mehr
The atmospheric INP measurements were performed with the newly developed cloud expansion chamber PINE (Portable Ice Nucleation Experiment), which is a mobile INP instrument optimized for automated and continuous INP measurements with a time resolution of about 6 min. In this work, PINE was operated at temperatures in the mixed-phase cloud regime between -15 °C and -30 °C. For an extension of the covered temperature range of the INP measurements to temperatures up to about -5 °C, INSEKT (Ice
Nucleation Spectrometer of the Karlsruhe Institute of Technology) analysis were performed. INSEKT is a well established method and was also used to validate the PINE measurements, but has the drawback of a poor time resolution of several hours or days.
Because PINE is a newly developed instrument, intercomparison and validation experiments with other INP instruments were performed at the AIDA (Aerosol Interaction and Dynamics in the Atmosphere) cloud simulation chamber. Intercomparsion experiments with 12 different INP instruments were carried out using 5 aerosol samples with different freezing properties in the mixed-phase cloud regime. Ambient INP concentrations are highly variable, so the study in this work included for the first time an intercomparison for a wide range of INP concentrations using the same aerosol sample. Independent of the probed INP concentration, the offset between different instruments was up to 4 orders of magnitude, similar to instrument offsets reported in Hiranuma et al. (2015) and DeMott et al. (2018). The reasons for these offsets remain unclear, but may be related to different ice formation modes or pathways the different instruments are sensitive for.
Ambient INP measurements at three rural locations and one mountain site were conducted continuously over a time period between 8 weeks and 5 months. Three of the campaigns included INP monitoring with PINE with a time resolution of about 6 min, and for all campaigns INSEKT INP measurements were performed with a time resolution between 4 h and 3 days. Generally, the INP concentration at all sampling sites varied between about 0.1 L^-1 and 300 L^-1 at a temperature of about -23 °C. The observed variation in the INP concentration measured with PINE in two rural environments showed enhanced positive correlations with the ambient air temperature, global radiation and planetary boundary layer height, and negative correlations with the humidity in time periods of a diurnal variation. In time periods without a diurnal variation of the INP concentration, no clear correlation with the analyzed variables was found. From this, one may conclude that the short-term variation of the INP concentration is caused by local effects, like farming activities or biological particles, which were not quantified by the measured parameters. Based on the measurements in the rural locations, a new parameterization was developed for calculating the INP concentration in models. The improved parameterization represents up to 30% more data points within a factor of 10 from the 1:1 line than previous parameterizations applied on the data set. Some existing parameterizations relate the predicted INP concentration to the number concentration of particles with a diameter larger than 0:5 μm. Also the INAS (Ice Nucleation Active Site) density approach implies that larger aerosol particles are more likely to induce ice formation. However, the impact of the size of an aerosol particle on its ice-nucleation activity is not yet well investigated. Laboratory experiments from this work with size-selected dust aerosol depicted that in some cases the freezing probability per particle surface area is independent of the size while in other cases smaller aerosol particles showed a higher freezing activity per surface area. This variable agreement was found to be potentially related to heat sensitive particles like organics or microbes, which may be more frequent at smaller particle sizes. Moreover, when conducting size-selective INP measurements in a rural environment it was observed that within a time period of about 36 h the freezing ability of the ambient aerosol changed from a size independent freezing to a higher freezing ability of larger aerosol particles. This change in the freezing ability was connected to a change in the air mass origin, which may have caused a change in the size-dependent aerosol composition. Therefore, future size-selective INP measurements at a high time resolution should be combined with size-resolved aerosol composition analysis at a similar time resolution.
Inspired by the reliable and precise operation of PINE, the new stationary and small cloud expansion chamber AIDAm (AIDAmini) was developed. AIDAm is located in the cold room of the AIDA facility and samples aerosol from the AIDA chamber for e.g. measuring steady changes in the ice nucleation properties of aerosol particles which may undergo chemical aging or other modifications during long-term aerosol simulation experiments. The validation of AIDAm showed a good agreement of AIDAm, AIDA and INSEKT measurements for selected operation conditions. It was also proofed that AIDAm is able to distinguish between homogeneous freezing of aqueous aerosol particles at temperatures lower than
-35 °C and the heterogeneous ice formation in the presence of INPs at the same temperature. With this new instrument, the temporal evolution of the ice nucleation ability of different aerosol samples during an hour-long coating procedure at conditions of the upper troposphere was investigated. Thereby, a 20 nm thick coating layer of sulfuric acid or secondary organic aerosol was established over 5 h, which fully suppressed or clearly shifted the ice activity of the pristine aerosol, respectively.
This thesis shows the high value of automated and continuous INP measurements with a high time resolution both in the laboratory and the field and quantifies changing ice nucleation properties of aerosol particles, which may be connected to variations of environmental or meteorological processes occurring on a range of time scales from minutes to days.