Mohtar, Zul Azmi
(2013)
Determination Of Coefficient For Pm10 Monitoring Using
An Optical Direct Reading Instrument And Beta Attenuation Monitor.
Masters thesis, Universiti Sains Malaysia.
Abstract
This research compares PM10 mass concentration record generated by two types of
aerosol monitoring devices, the Optical Direct Reading Monitor (ODRM) and the
BAM1020, were collocated outdoors at Perai and Seberang Jaya. ODRM records
were adjusted to predict Beta Attenuation Monitor (BAM) with a correction factor k
derived from co-located BAM and ODRM at a research site to make data comparable
in real terms. Predicted BAM records were obtained using quantile regression.
Performance indicator such as Coefficient of Determination (R2), Index of
Agreement (IA), Normalised Absolute Error (NAE) and Root Mean Square Error
(RMSE) were used to measure the accuracy of the predicted BAM and actual BAM.
The k was successfully identified for Five Groups of Percentile and Seven Groups of
Percentile. For both methods, monitoring records were treated using two techniques
and named as Clean Record and Original Record. Clean Records removed the first
two days records and concentrations lower than 10 μg/m³. For Original Record, only
the first two days (48 hours) records were removed. Positive significant correlation
was found between PM10 and temperature at all station using ODRM and BAM.
However, for correlation between PM10 and RH, all stations showed negative
correlation except Perai station using ODRM (0.901). The equation model used for
predicted BAM is . pred BAM kx
which is . is constant value and x is observed
ODRM. Five Groups of Percentile using Clean Record is the best for both stations.
R2 value for Five Groups of Percentile for Perai is 0.9170 and 0.9422 for Seberang
Jaya compared with Seven Groups of Percentiles that is 0.8795 for Perai and 0.8402
for Seberang Jaya. Hence, simple optical direct reading instrument can be used as a
surrogate to measure PM10 following statistical calibration by using quantile
regression.
Actions (login required)
 |
View Item |
