| Summary: | 碩士 === 中興大學 === 環境工程學系所 === 94 === My research data source is supplied from air quality monitoring station and weather station that Central Weather Bureau set up within the border between Taichung and Nantou area, analyze and process this data in seven years from 1997 to 2003. In order to understand wheather every pollutant have influenced the change of ozone in Nantou County, the analysis method I use is Data Mining.The result of Data Mining that every method gets is: (1)Decision Tree : When between eleven o’clock in the morning and five o’clock in the afternoon ,NO concentration is less than 2.95ppb, PM10 concentration is more than 56.85ug/m3 , NO concentration is less than 0.95ppb, SO2 concentration is under 2.15ppb, and the atmospheric temperature of the monitoring station is up to more than 25.95 degrees, then the average of the ozone concentration will be up to 80.2ppb. (2) Regression Model :In March,May,September and October between eleven o’clock in the morning and five o''clock in the afternoon,rainfall rareness, rainfall time are short, the sunshine is long, in this situation,ozone value is relatively high at this moment; When NOx, NO, NO2, THC, NMHC concentration increase, ozone value will drop.(5) Cluster Analysis: The relative humidity is low, the wind speed is on the low side, there is little rainfall, visibility is good, the cloud is relatively low in height, the insolation amount is relatively strong, dew point temperature and atmospheric temperature is relatively high, NO, NO2, NOx is on the low side, in Bamboo Mountain and Nantou Monitoring station, happened in August, September, November,at noon, two o’clock ,three o’clock and six o’clock in the afternoon, ozone value situation will be on the high side. (4) Neural NetWork: Use the data for seven years via training and testing, then produce neural equation, take the place of the pollutant concentration which enter 2005 and contrast observation value and simulation value,compare the change of the trending with the results produced from regression equation.(5) Discrimination:The most important factors that affect ozene concentration better or worse are NO, NO2, NOx and CO.
|
|---|
