In this particular case what you are looking for are pollutants that reside in the boundary layer (BL). The BL height (depth of mixing layer) varies continuously with the time of the day. This happens because the BL height depends directly on the surface temperature and the amount of insolation. So, in the night and early morning you could easily have a very shallow BL such as some 200m or less. However, as the sun heats up the surface, the temperature inversion breaks and the BL starts to grow. Depending on the terrain and the season, you could have a noon BL as deep as 2000m. Now, it is important to mention that the HYSPLIT model uses meteorological fields that are outputs from global models. This implies that the larger scale meteorological features are very well captured but the smaller scale local motions are not. So, the caveat for your particular application is that the HYSPLIT model is not good to find back trajectories in very shallow layers adjacent to the surface (~200m or less). This is something that you need to keep in mind. From my perspective, tracking where the air comes from for small scales of motion is more suitably done by using local meteorological observations. I would suggest finding met stations around the area that you are interested in and looking at the wind speed and direction for the days you need. It is not unusual to find out that there are seasonal winds and sometimes sustained tendencies for directions for certain hours and months within a particular location. Hope this helps. Good luck!

Dear Ms.Cazorla, Thank you for your very informative suggestion. :)

I completely agree with Maria.

But you should be aware of what you are analyzing. Dew is condensation of water vapour onto a surface. so per definition dew is rather pure and only diffusion processes and deposition/sedimentation is transporting pollutants onto the collection surface. If this surface is exposed for a long time dry deposition of aerosol particles will determine the composition of collected dew water. If there are fog droplets present  wet deposition will transport pollutants ont the collection surface too (concentration in fog droplets are about 100 times higher than in rain drops).

So the concentration you are measuring is a mix of dry deposition during exposure time of the collection surface with no dew formation + dry deposition during exposure time with dew formation+wet deposition of fog droplets (if fog is also present) + flux of gases onto the wet surface (respect Henry constants of gases) and chemical reactions taking place in the liquid phase . Maybe there is an enhanced deposition of particles caused by the diffusion flux  of the water vapour. For very small particles this also may prohibit a depositon to some extend as the flux of water molecules to the collection surface causes a counter flux of other molecules to the opposit direction of the condensation flux. 

The first two answers raise some salient points with which I concur. Something else to be borne in mind is the chemical composition of air close to the ground; as well as varying with the PBL meteorology as described in the first answer, there are many shortlived gases, some of which are either or both reactive and soluble. Ammonia is prominent among these, depending on location and local biota, it can have very large fluxes from and to the surface and its biota. There is often also a rich variety of organic compounds, viruses and bacteria which are very far from being well characterised or understood. In principle and in practice, they can affect the surface characteristics of water droplets, see for example Donaldson & Vaida, 2006. Chemical Reviews, 106(4), 1445-1461. Finally, equilibrium is a bad assumption in most cases of atmospheric phase changes.

Thank you Mr. karl and Mr.Adrian, I am from Chemistry background (M.S. student), so although Ms. Maria gave very good suggestion but i had to do some study to understand her point. But both of your answer is crystal clear to me, it seems i am hearing from a chemist :) . Mr. Adrian you are right, i have observed that ammonia is a dominant chemical species in dew water (about 50-1800 microequivalent per litre reported from different part of the world). and there is report of studying biological characteristic of dew water ( Chemical and biological characteristics of dew and rain water in an urban coastal area (Bordeaux, France), Atmospheric Environment 40 (2006) 3710–3723). This discussion was very helpful and increase my understanding about dew.  

http://www.sciencedirect.com/science/article/pii/S1352231006002688

thanks for the feedback. When you are talking of ammonia in dew water you are most probably talking on the ammonium ion. Please be aware that ammoniumsulfate is a major constituent of aerosol particles and significantly influences the pH-value. As Adrian Tuck mentioned short lived gaseous constituents (i.e radicals formed by photochemistry) are present especially during daytime. H2O2 may also be formed in the liquid sample if exposed to sunlight. Ongoing chemical reactions may change the pH-vlue of the sample rather quickly .

And by the way I am a meteorologist. :)

By training I'm both a physical chemist (Cambridge, 1969) and a meteorologist (Meteorological Office College, 1976).  From long experience with both Eulerian and Lagrangian approaches to atmospheric transport, I'll offer the following: trajectories calculated from analyses (such as Hysplit) are really Lagrangian sampling of an Eulerian system. Theoretically, "pure" Lagrangian approaches are inadequate, because they are solely about organized flow, while in reality dissipation (decorrelation of structure in fluctuations) never quits. There are attempts to deal with this, pioneered by Daniel S McKenna and co-workers at KFA Juelich. The straight Eulerian flux approach in numerical models does not treat dissipation in a physically realistic way. Finally, it's my opinion that the way Hysplit deals with flow over topography can lead to large errors, something particularly severe in the unique conditions of poleward flow encountering the Antarctic ice cap, but that's another story.

Mr. Karl : Sorry for the silly mistake, I was actually talking about Ammonium ion (check the unit m.eq/L). The dew sampler was placed at night and sample was collected before sun rise to minimize evaporation by sun. and its a pleasure to meet a meteorologist, who has very good understanding of Chemistry. :)                                       Mr. Adrian: The fact is that, i dont have very good knowledge about modeling.  another problem is that there is no nearby weather station, from where i may get wind direction (as Maria suggested) during my sampling time. so, HYSPLIT is the only option i have, because its easy to construct. I just go to the NOAA arl website, put the parameter and get my trajectory analysis (although there is chance of large error).                                     Anyway, it was nice to know about you. :)