An additional study the trend of the two parameters is to use the AOT value extracted from the AOT
product available within one month at the pixel where a interested measurement station is located to plot a
graph to see the its monthly temporal distribution. Similarly, hourly PM10 concentration value recorded
within a period of the same month at the same measurement station can be used to plot a graph to see
observe the trend of its temporal distribution.
2.2 Generation of Estimated PM10 Concentration Maps for Upper Northern of Thailand
In previous section, we have analyzed the trend of yearly and monthly distribution of MODIS AOT
and PM10 concentration obtained for the location of 12 measurement stations in the study area in the
period of 2007 to 2011, and obtained fairly relationship, as the correlation coefficient is 0.32, with the
linear regression equation of y = 57.09 x + 70.93, where, x is AOT value in digital number (DN), it
should be multiplied by a scale factor of 0.001 to be AOT physical parameter, and then can be converted
to PM10 value. We apply this equation to convert AOT to PM10, even though the correlation is fairly small
(correlation = 0.32), but since it was generated from the data collected from 12 stations located in the
study area within the period of 4 years from 2007 to 2010. Figure 2 shows the quick look image of the
MODIS AOT map (a) and the estimated PM10 concentration (b) generated from the MODIS AOT map
for Southeast Asia region.