Southern Alaska as a source of atmospheric mineral dust and ice-nucleating particles

Abstract

Ice-nucleating particles (INPs) influence cloud radiative properties and climate; however, INP sources and concentrations are poorly constrained, particularly in high-latitude regions. Southern Alaska is a known source of high-latitude dust, but its contribution to atmospheric mineral dust and INP concentrations has not been quantified. We show that glacial dust collected in southern Alaska is an effective ice-nucleating material under conditions relevant for mixed-phase clouds and is more active than low-latitude dust because of a biological component that enhances its activity. We use dispersion modeling to show that this source contributes to the regional INP population and that the dust emitted is transported over a broad area of North America, reaching altitudes where it could cause cloud glaciation. Our results highlight the importance of quantifying emissions and ice-nucleating characteristics of high-latitude dusts and suggest that the ice-nucleating ability of emitted dust in these regions should be represented in models using different parametrizations to low-latitude dust.

Fig. 1.. True color image of the Copper River Delta.

Derived from Landsat 8 Collection 2 Tier 1 calibrated top-of-atmosphere reflectance of a dust event on 31 October 2020.

Fig. 2.. Size-resolved ice-nucleating activity of Copper River Valley dust.

(A) Fraction frozen for all samples. (B) INP concentration per standard liter of air for all samples. Different colors represent each of the four collection stages, and different symbols distinguish different sampling days. The mean and SD of handling blanks shown in (A) represent the background INP activity used in calculating the error shown by error bars in (B).

Fig. 3.. Heat tests for protein-based biological ice-nucleating entities.

Fraction frozen before heating (blue) and after heating (red) of each sample along with box plots for each stage. The box plots include all of the samples; each box represents the 25th and 75th percentiles, and the whiskers cover the full spread of the data. The median freezing temperature, ΔT₅₀, is shown with a black line.

Fig. 4.. Size-resolved active site density (n_s) of Copper River samples.

(A) Comparison of n_s for Copper River dust to n_s parametrizations of ice-active minerals from the work of Harrison et al. (39) scaled to the mineral content of our bulk sample (from XRD analysis): 9% K-feldspar, 15% quartz, and 58% albite. (B) Comparison of n_s for Copper River dust to airborne dust samples from Iceland (52) and Svalbard (48) and sediment samples from Iceland (42) and the Yukon (22). (C) Comparison of the n_s for Copper River dust before and after heating, as well as to the same minerals as in (A).

Fig. 5.. FLEXPART results from a 10-day simulation of a dust event at the Copper River Delta.

(A to D) Transport of 15 kt of dust that was released over 4 days, starting 14 October 2019. The total column dust mass [0 to 10,000 meters above ground level (magl)] for 48, 96, 144, and 192 hours after the start of the emission period is shown. (E) Vertical profile of the mean position of released particles (blue line), the boundary layer height (dashed line), and topography (green) along this trajectory.

Fig. 6.. Case study of dust and INP concentrations after 60 hours.

(A) Total dust mass integrated over a 0- to 5000-m column. (B) Vertical transect of dust concentrations along the red line shown in (A). Isotherms (from ERA5 reanalysis data) are shown in gray, and the topography (from FLEXPART output) is shown in green. (C) Vertical profile of mean dust concentration calculated within the red box shown in (B). Dashed lines represent isotherms of mean temperature in the same region. (D) Vertical profile of ambient INP concentration calculated using our n_s parametrization, dust concentrations shown in (C), and the mean temperature.

Fig. 7.. The Copper River Delta showing sampling locations A and B, Cordova, and the Copper River Highway.

Background image derived from Landsat 8 Collection 2 Tier 1 calibrated top-of-atmosphere reflectance.

Fig. 8.. Size-resolved dust sampling in the Copper River Valley.

(A) Sioutas Personal Cascade Impactor. Sampled air passes through accelerator plates A to D in turn and particles above the cut-off size for each plate (A, >2.5 μm; B, 1.0 μm; C, 0.5 μm; and D, 0.25 μm) are collected on to the corresponding collection plate. Collection plates (inset) consist of a 25-mm collection substrate, filter retainer, and nitrile O-ring to maintain an airtight seal. (B) Cascade impactor and optical particle counter deployed in the Copper River Valley on a tripod.