Air-passive sampling – recalculation

The GENASIS model for recalculation recalculates the passive sampling results using an exponential regression:

where the constants k_t, k_w and k₀ were estimated by using 8-year long time series of passive and active POPs monitoring at the Kosetice observatory.

Recalculation models

There are two completely different metodologies how to compute real concentration from the passive sampler data. The Genasis model is based on 8-year simultaneous measurements with active and passive samplers in Kosetice. In addition to the measurement itself, temperature and wind speed were recorded and consequently a linear model was created, putting all these quantities into context:

where c_passive denotes the result of passive sampling in mass/filter/day units and k₀, k_t and k_w denote the constants to be estimated by the model. Once the (compound-specific) constants were estimated, they were used in the same manner to determine the concentration. Similar constants were also computed for temperature dependent model and also for model which is independent on the external conditions (the less precise one).

Entirly different methodology is used inside the GAPS model, described in detail in the article by Harner et al. "Calibration and application of PUF disk passive air samplers for tracking polycyclic aromatic compounds (PACs)" (see reference below).

There are only a few options how to customize the function genpastoact. The necessary inputs are name of the pollutants to recalculation (there are 30 different PAHs, PCBs and OCPs included in the function), length of sampling period in days (28 days in our example) mean temperature and wind speed during the measurement and one of the sampling methods.

In this case, we load the data on temperature and wind first and then chose the GAPS model for windy sites (method=”gaps7”) and GENASIS model taking into account both temperature and wind (method=”tempwind”):

temp<-data[,"temp"]

wind<-data[,"wind"]

genpastoact(flu,days=28,method="gaps7",temp=temp,pollutant="fluorene")

genpastoact(dde,days=28,method="gaps7",temp=temp,wind=wind,pollutant="p,p'-DDE")

The result we obtained consists of the new vector of class numeric, containing the real ambient air concentrations (ng/m³):

genpastoact(flu,days=28,method="gaps7",temp=temp,pollutant="fluorene")

[1]   4.04787796  4.91295535  1.60894533  0.92050611  0.74699532  0.47183319
[7]   0.52785747  0.05222585  0.39880447  1.22592378  3.56051909 13.84909273
[13] 12.16772900  6.09804768] 4.33875193  3.55407707  1.62991660  0.50627749
[19]  1.12833293  1.25132221  0.93904797  0.88376874  0.53362438  3.34841542
[25] 10.33449160  8.35420254  6.37613494  8.78225359  3.94410374  2.59871404
[31]  1.00372990  0.81497059  0.80559454  0.93024679  0.49010667  0.66782373
[37]  1.71191966  3.00132129  4.86185691  1.51589306  2.63245253  1.90262024
[43]  1.26953149  0.83449264  0.42284469  0.34071244  0.35063701  0.64492357
[49]  0.98386671  2.78422366  3.07584742  5.64268645

genpastoact(dde,days=28,method="gaps7",temp=temp,wind=wind,pollutant="p,p'-DDE")

[1]  0.035248697 0.024945025 0.030262030 0.038093072 0.016182639 0.020823315
[7]  0.020896702 0.021748722 0.032199289 0.029597355 0.086565110 0.080450148
[13] 0.080265423 0.020689542 0.028053793 0.016806546 0.042687961 0.037396446
[19] 0.026142054 0.030823881 0.030466647 0.034268241 0.032528435 0.036388040
[25] 0.050449350 0.027045932 0.016437744 0.008371257 0.012458038 0.025441644
[31] 0.041012754 0.055719248 0.057369055 0.042927993 0.028303249 0.041343837
[37] 0.043907564 0.048967183 0.020262896 0.029340806 0.015634150 0.025794940
[43] 0.013401002 0.021336624 0.026987077 0.012603440 0.016336436 0.044402436
[49] 0.045631194 0.018464170 0.015643061 0.013893827