particular day.
This set of data is also analyzed with linear regression. A linear relationship is found between logarithms of bacterial
concentration and temperature. For samples collected in the nonventilated room, the relationship is
ln(C) = -39.7 + 0.65 x T (eq. 3-3), where 66.1 °F < T < 70.4 ^F, and 15.9% < RH < 21.4%, but the equation is confounded by relative humidity (Figure 8, 9, and 10.) . The maximum concentration occurs at a relative humidity of 19.5%.
Table 8. Variance component for samples collected in the room without ventilation (data transformed to log scale).
Variance component Estimated variance
Day-
Run within one day
Pair (inter-paired) Error (intra-paired) 0.2436 0.1655 0.01237 0.1146
Temporal variance component = Day component + Run component
Spatial variation, inter-paired variation, is tested
plotted with time (Fig.4). Average bacterial concentration
decayed over time for these data. Samples collected earlier on a particular day are more variable than those collected
later on the same day.
2. Samples collected in the outdoor environment
Samples were collected on the roof of the Annex building (Fig.2). A total of 39 samples for each sampler were
collected. Four sets of data are respectively assigned C^, Cb, Cc, and C^ to indicate that samples were collected with samplers A, B, C, and D. The distance (d) was 0 cm between samplers A and B, 100 cm between samplers A-B and C, 200 between samplers C and D, and 224 cm between samplers A-B
and D.
Data is analyzed with the t test and the distribution- free sign test. The results show that concentrations are equal for bacteria collected with samplers A, B, C, and D. In order to study the influence of distance on spatial
variations, semivariogram values are also computed for this set of data. The semivariogram values (Table 9) do not
increase with the incremental distance between the two
samplers, nor do they increase with the mean concentrations
Table 9. The Comparison of Semivariogram Values for Samples Collected Simultaneously by four Samplers on the Roof of the
Annex Building.
Distance Mean of Number of Semivariogram between bacteria samples values {*)
two concentration ' samplers (CFU/m^) (cm)
1 °
90.6 39725 1
1 100
91.4 39 587 200 82.8 39 64 61 224
86.3 39 527 58ͫ
collected at each distance between samplers. Based on the semivariogram values, spatial variation in the outdoor
environment was not dependent on distances between samplers
within this small scale, 224 cm.
Polynomial regression analysis has been considered to determine the relationship between CFU concentration as a function -of temperature and relative humidity. The results show that the bacterial concentration has a quadratic
relationship with temperature (T°F) and relative humidity
(RH%). The relationship is
C = -148 + 7.13 * T + 1.82 * RH + 0.135 * T * RH - 0.096
* T^ - 0.199 * RH^ (eq. 3-4).
R^ = 0.6 and the Pearson correlation coefficient =
0.8054 at 33*F < T < 67. 5*F and 7.5% < RH < 58%, where C is
the average corrected colony counts of four sets of samples. This data is also plotted with time to represent
temporal variation (Figure 11.). Bacterial concentrations
do not decay over time.
3. Samples collected in the ventilated room
Samples were collected with four samplers at three
locations in room S-250 (Fig.3). A total of thirty samples was collected. C^ is the bacterial concentration collected with sampler A at the door of the room, C^ is the bacterial
y^^t^^^^j^-ir^rspfc'-
concentration collected with sampler B in the middle of the room, and Cc is the bacterial concentration collected with sampler C deep inside of the room. Data is tested using the distribution-free sign test. Results show that the highest bacterial concentration is collected at the middle of the room, the second highest concentration is collected at the deep inside of the room, and the least concentration is
collected at the door of the room (Table 10).
This set of data is also analyzed using linear and polynomial regression. No statistically significant
relationship is achieved between bacterial concentrations and temperature and relative humidity. Bacterial
concentrations are also plotted with time to represent
temporal variation (Figure 12). Bacterial concentration do not decrease over time but vary randomly.
4. Samples collected during the day in the nonventilated
room
Samples were collected with the same sampler arrangement for variance component analysis (Figure 1). The distance
Table 10. Distribution-free sign test
in the ventilated room.
for samples collected
Sample II ^°- Cone. A (CFU/M^) Cone. B (CFU/M^) Conc.C CFU/M^) Sign (A - B) Sign (C - B) Sign (C-A)
1 ͣ"ͣ
46 86 7 0 00 1
1 ^
36 50 57 0 10 1
1 ^
36 61 46 0 00 1
1 ^
36 46 29 0 01 1
1 ^
4 36 43 0 10 II
1 ^
21 36 36 00 II
1 "7
25 50 36 0 00 II
1 ^
32 46 39 0 00 II
1 ^
32 64 46 0 00 II
1 -^^
25 68 39 0 00 1
11 54 46 68 1 10 II
II -^^
21 32 25 0 00 1
II 13
32 61 32 0 0II 14
54 64 11 0 0 11 ^^
68 86 68 0 00 1
1 ^^
39 68 43 0 00 II
1 ͣ'•'^
32 43 36 0 0 01 -^^
50 86 75 0 00 II
1 -^^
129 139 61 0 01 II
1 ^°
54 118 79 0 00 1
1 ^-^
57 96 64 0 0 01 ^^
21 71 46 0 0 0 23 43 64 61 0 0 0 24 29 43 82 0 1 0r 25
32 68 61 0 00 1
611 26
25 50 25 0 01 ^"^
32 46 46 0 01 28
43 79 57 0 0 01 29
71 146 118 0 0 01 30
107 139 111 0 00 II
1 Sum
1 43 1
62between two pairs of samplers was also 470 cm. A total of thirty samples was collected on three consecutive days. Data is analyzed using the distribution-free sign test (Table 11.). Spatial variation is statistically significant, and bacterial concentrations collected with samplers close to the door are higher than those collected with samplers far away from the door. Bacterial concentrations are also plotted with time to study temporal variation.
Concentration significantly decreases after a sampling period of seven hours. The period of the seven hour sampling time was usually from 10:00 am to 5:00 pm.
D. Respirable and irrespirable counts
After the total counts are analyzed to study temporal and spatial variations, each sample is separated into
respirable and irrespirable counts, then they are
respectively analyzed to study the relationship between