Laserfiche WebLink
observations. Calculations were performed using the NOAA Air Resources HYbrid <br />Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model available at <br />http://www.arl.noaa.gov . Model documentation (Draxler and Hess, 2004) is available at <br />http://www.arl.noaa.gov/data/web/models/hysplit4/win95/arl-224.pdf . For the 2004 data, <br />back-trajectories were performed using the NCAR/NCEP (National Center for <br />Environmental Prediction) re-analysis data as the meteorological basis for the <br />computation. This data set was not available this fall for analysis of 2005 trajectories. <br />Instead, two sets of trajectories were calculated. For each set, the back-trajectories of the <br />parcel arriving at 3 m AGL (roughly the sampling height) and 500 m AGL (roughly the <br />upper portion of the planetary boundary layer at mid-day) were calculated. One set was <br />based on the NCEP Eta numerical weather prediction model Data Assimilation System <br />(EDAS) initial analyses and the other was based on initial analyses from the Global Data <br />Assimilation System (GDAS) associated with NCEP GFS (Global Forecast System <br />weather prediction model). The results typically varied for the different heights within <br />each set, and between sets. Examples for 3 July, when there is reasonably good <br />agreement, are shown in Figures 5 and 6. The upper trajectory is in blue, the lower one in <br />red. If precipitation occurred along the trajectory, this is indicated by the red and blue <br />traces in the panel below each map. <br />Agreement was not always so good. Figures 7 and 8 show an example of poor agreement <br />between trajectories at different height and based on different meteorological analyses. <br />Following the analysis procedure adopted in Detwiler et al. (2005) for the 2004 data, the <br />3-day air back-trajectory arriving at Stanley on each sampling day was classified into a <br />general compass direction and distance. For the 2005 data, it was additionally possible to <br />look at the differences between the two sets of calculations based on the two different <br />meteorological data sets. On most days there was rough agreement in trajectory character <br />and general compass direction. Based on this analysis we identified a few trajectory <br />groups that were identified consistently with either high or low CCN concentration. <br />As in Detwiler et al. (2005), trajectories arriving from the general direction of the <br />northwest or north were generally characterized by low CCN concentrations. Examples <br />of days with such trajectories are 3 and 5 June. We attribute the low CCN concentrations <br />to lack of CCN sources over the less-developed regions of central and northern North <br />America. <br />Also, similar to the findings from Detwiler et al. (2005), trajectories from the west, <br />particularly those originating as far west as western Idaho, Washington, and Oregon, <br />were sometimes characterized by high CCN concentrations, as on 15 July. Other days <br />with similar trajectories sometimes did not have such high CCN concentrations. It is <br />possible that localized and intermittent forest fire activity to the west could have been the <br />source for occasional relatively high CCN concentrations. <br />8 <br />