Abstract. Produced by the incomplete combustion of fossil fuel and biomass, black carbon (BC) contributes to Arctic warming by reducing snow albedo and thus triggering a snow-albedo feedback leading to increased snowmelt. Therefore, it is of high importance to assess past BC emissions to better understand and constrain their role. However, only a few long-term BC records are available from the Arctic, mainly originating from Greenland ice cores. Here, we present the first long-term and high-resolution refractory black carbon (rBC) record from Svalbard, derived from the analysis of two ice cores drilled at the Lomonosovfonna ice field in 2009 (LF-09) and 2011 (LF-11) and covering 800 years of atmospheric emissions. Our results show that rBC concentrations strongly increased from 1860 on due to anthropogenic emissions and reached two maxima, at the end of the 19th century and in the middle of the 20th century. No increase in rBC concentrations during the last decades was observed, which is corroborated by atmospheric measurements elsewhere in the Arctic but contradicts a previous study from another ice core from Svalbard. While melting may affect BC concentrations during periods of high temperatures, rBC concentrations remain well preserved prior to the 20th century due to lower temperatures inducing little melt. Therefore, the preindustrial rBC record (before 1800), along with ammonium (NH4+), formate (HCOO−) and specific organic markers (vanillic acid, VA, and p-hydroxybenzoic acid, p-HBA), was used as a proxy for biomass burning. Despite numerous single events, no long-term trend was observed over the time period 1222–1800 for rBC and NH4+. In contrast, formate, VA, and p-HBA experience multi-decadal peaks reflecting periods of enhanced biomass burning. Most of the background variations and single peak events are corroborated by other ice core records from Greenland and Siberia. We suggest that the paleofire record from the LF ice core primarily reflects biomass burning episodes from northern Eurasia, induced by decadal-scale climatic variations.

The snowpack acts as a sink for atmospheric reactive nitrogen, but several postdeposition pathways have been reported to alter the concentration and isotopic composition of snow nitrate with implications for atmospheric boundary layer chemistry, ice core records, and terrestrial ecology following snow melt. Careful daily sampling of surface snow during winter (11-15 February 2010) and springtime (9 April to 5 May 2010) near Ny-Ålesund, Svalbard reveals a complex pattern of processes within the snowpack. Dry deposition was found to dominate over postdeposition losses, with a net nitrate deposition rate of (0.6 ± 0.2) μmol m À2 d À1 to homogeneous surface snow. At Ny-Ålesund, such surface dry deposition can either solely result from long-range atmospheric transport of oxidized nitrogen or include the redeposition of photolytic/bacterial emission originating from deeper snow layers. Our data further confirm that polar basin air masses bring 15 N-depleted nitrate to Svalbard, while high nitrate δ( 18 O) values only occur in connection with ozone-depleted air, and show that these signatures are reflected in the deposited nitrate. Such ozone-depleted air is attributed to active halogen chemistry in the air masses advected to the site. However, here the Ny-Ålesund surface snow was shown to have an active role in the halogen dynamics for this region, as indicated by declining bromide concentrations and increasing nitrate δ( 18 O), during high BrO (low-ozone) events. The data also indicate that the snowpack BrO-NO x cycling continued in postevent periods, when ambient ozone and BrO levels recovered.

No abstract

ABSTRACT. Spatial heterogeneity of snow and firn properties on glaciers introduces uncertainty in interpretation of point and profile observations and complicates modelling of meltwater percolation and runoff. Here we present a study of the temporal and spatial dynamics of firn density and stratigraphy at the plot-scale (≈10 m × 10 m × 10 m) repeated annually during 2012-14 at the Lomonosovfonna icefield, Svalbard. Results from cores, video inspections in boreholes and radar grid surveys are compared. Ice layers 0.1-50 cm thick comprised ≈8% of the borehole length. Most of them are 1-3 cm thick and could not be traced between boreholes separated by 3 m. Large lateral variability of firn structure affects representativeness of observations in single holes and calls for repeated studies in multiple points to derive a representative stratigraphy signal. Radar reflections are poorly correlated with ice layers in individual boreholes. However, the match between the high amplitude peaks in the grid-averaged radar signal and horizons of preferential ice layer formation revealed by averaging the video surveys over multiple boreholes is higher. These horizons are interpreted as buried firn layers previously exposed to meltfreeze or wind-driven densification and several of them are consistently recovered throughout three field campaigns.