A trajectory climatology for Svalbard;: investigating how atmospheric flow patterns influence observed tracer concentrations

A 10-year climatology of long-range atmospheric transport to Ny-(A) over circle lesund, Svalbard (78.9degreesN, 11.9degreesE) is developed using three-dimensional 5-day back-trajectories. We calculate trajectories arriving twice daily at 950, 850 and 750 hPa during 1992-2001, using European Centre for Medium-Range Weather Forecasts (ECMWF) analyzed wind, fields. Cluster analysis is used to classify the trajectories into distinct transport patterns. The clustering procedure is performed on the whole 10-year set of trajectories, to study both year-to-year and mouth-to-mouth variability in the synoptic-scale atmospheric circulation. We identify eight major transport patterns to Ny-(A) over circle lesund, which we find to be consistent with mean-pressure charts of the Arctic region. The distribution of trajectories between these flows is similar for all years during the 10-year period. However, there are seasonal differences in when different clusters are most prevalent. The calculated clusters provide an indication of source regions and transport pathways influencing Svalbard at different times of the year. Such information is valuable for interpreting measured time-series of trace gases and aerosols and could serve as guidance for formulating sampling strategies. We compare the trajectory clusters to CO2 measurements to study to what degree different atmospheric flow patterns influence the variability of the atmospheric CO2. Overall we see a linkage between CO2 concentration and the large-scale circulation. For instance, in connection with transport over Europe and Siberia during winter, high CO2 mixing ratios are observed, whereas trajectories originating from the Atlantic are associated with low CO2 concentrations. However, during some periods and for some individual trajectories we see no conclusive linkage between variability in atmospheric CO2 and transport. This can be due to a combination of the complex structure Of CO2 sources and sinks and its relatively long atmospheric turn-over time. CO2 and Rn-222 mixing ratios are calculated using the three-dimensional transport model MATCH to further illustrate these characteristics of CO2. (C) 2003 Elsevier Ltd. All rights reserved.