CHAPTER 6
CONCLUSION AND FUTURE WORK
6.1 CONCLUSION
Monitoring air pollution for specific region using geospatial technology can be
used in many fields and purposes. Monitoring air pollution for specific region is an
essential part of the disaster management system. Its results may assist disaster
management for the design of future smart air pollution monitoring system and control
the effects caused faster than earlier system. Pollution free zone may be improved by
applying immediate measures against the pollutants detected.
More than the study area, and the distribution of spatial data for different
concentrations of PM10. Based on exploratory data analysis, and distribution of
concentrations of PM10 tends positively due to higher concentrations then occurred
limited values in the part of the region. A range of concentrations of PM10 between
0.20μg / m3 and 455.0μg / m3.
The link has been examined only on measurements in situ observations PM10
since the height of land cover types are fixed in time compared to the daily variation of
the concentration of PM10. There are about 4 consecutive days observed a significant
correlation. Mutual link between the stations are relatively lower than the link cars.
Data integrity is a major concern when it comes to dealing with different types
of data from different sources. Search deal with remote sensing products: DEM map land
cover, and on-site measurements and data CTM network results. To avoid uncertainty
comes from the reduction of remote sensing products, which represents a network data
CTM as a point of information has been rising and land cover data network
corresponding basis points set. However, one can argue that this method is something
very naive to do and which is not considered to raise issues in the integration.
It has developed a spatial separation model by adding the concentration of
PM10 Functions variogram spatial data set. Results CTM and CORINE land cover
modeling in the explanatory variables were included. Known function of spatial and
variogram as a function of distance HC spatial data and Hizb ut-Tahrir, respectively. It
has been implemented Kriging approach based on this estimate variogram. However, this
modeling, kriging requires a full-time series at each site.
Map is created using the excesses of the concentrations of PM10 indicator
kriging. Below are assigned the notes a certain threshold to 1 and is set to zero notes
exceeded. Based on the style variogram spatial index, values were obtained to predict the
probability. Most importantly, the calculated threshold probability value (0697) to create
a map of daily excesses in January is critical, where these marginal values are calculated
from all the data in January.
An area that has a high probability of exceeding the concentrations of PM10
vary from one region to another. However, a map showing the probability that the
probability of exceeding likely to be higher in the eastern part of the study area, and
communicate with land cover, with an area covered by artificial surfaces that have a
greater likelihood than others.
Measurement errors were an integral part of the observations that cause errors in
modeling and analysis processes. Similarly, the conclusion of this research is not without
errors. Therefore, after the investigation of different methods of noise and remove the
error, in this chapter it is found an enhanced Savitzky-Golay to be more convenient than
other methods such as wavelet analysis. This filter can not compensate alone for the loss
of data. Therefore, this research uses texture with the functions of a slice orthogonal
polynomials Model. The former being the right lost daily and weekly data and last for
missing monthly data. And the extent of the impact of these methods to improve the
accuracy of prediction of nearly a model of10-15%. Two data sets were used in this
research, which is fine-tuning or not refine time series of pollutants difference. In both
cases, it is the best way to predict air pollution NARX way. When refining a time series
of pollutants, about40% increase in R2 and 94% decrease in RMSEhappen or occur. In
view of the previous studies, this research is the slightest error in predicting air pollution
model.
Directed times the maximum prediction also, which achieved great accuracy
during 7 days. The results of this search are only for PM10 and PM2.5 pollutants as the
main source of air pollution in Tiruchirappalli. The applicable data collection also from
the stations to measure air pollution on a daily basis. finally, using a genetic algorithm,
Effective parameters have been identified in the prediction of air pollution data for the
day of the week, a week of the month, terrain, wind direction, maximum temperature, and
the value of pollutants for two of the nearest neighbor air pollution stations selected.
Air pollution and hazardous chemicals play a role in human health and plants. Chemical
effects of air pollution on health is very complex and there are various sources and
individual impacts shift from one to the next. It is assessed air quality encompassing from
various pieces of the industrial zone and Chennai . The air quality is difference analysis
and classification on the basis of the air quality index table. Before the data and data
processing can be further processed. Since the government adopted a series of measures
to combat this problem. It will be taken to predict chemical air pollution in urban and
industrial areas in Chennai research study using data mining forward with the retrieved
data from monitoring stations state of Tamil Nadu, India.
The data collected from pollution from different monitoring stations in
Alandhur, IIT, Manali, residential areas such as Annanagar, monasteries, T.Nagar,
Nongambacam, Kilpauk and Valluvarkottam in Chennai. Industrial zones such as
Sriperumbudur data, Alwar Thirunagar and Manali are also collected and converted to a
CSV file format.
6.2 FUTURE WORK
The proposed system worked only with the local air pollution data. Further by
enhancing Internet of things technology with geospatial technology, all globally
communicating devices can operate with revisited protocols that can interpret global data.
This can expose new technology that enables sophisticated access to human against
disaster management for their society. Also, the low level primitives of the air pollution
are addressed in this work that restricted some results can be enhanced by accepting
dynamic input from advanced new Geocomputing methods like merging Mobile
computing and Internet of things.
