INFLUENCES OF METEOROLOGICAL PARAMETERS AND OZONE PRECURSORS TO NEURAL NETWORK MODEL IN OZONE TROPOSPHERIC CONCENTRATION (STUDY IN BANDUNG CITY)

Dyah Lukita Sari (Pusat Penelitian dan Pengembangan, Badan Meteorologi Klimatologi dan Geofisika Jln. Angkasa I No. 2 Jakarta)
R. Driejana (Kelompok Penelitian Manajemen Udara dan Limbah, Departemen Teknik Lingkungan, Institut Teknologi Bandung)

Widyariset
Vol 14, No 3 (2011): Widyariset

As an air pollutant, the fate of ozone is determined by the movement and chemistry of atmospheric dynamics and the precursors of ozone are emitted from the source. In that case the effect of each of meteorological param- eters and ozone precursors need to be known so that the ozone concentration can be predicted. In this research the concentration of ozone precursors (NO and NO2), temperature, solar global radiation, humidity, and wind speed as meteorology parameters were used as input in artifi cial neural network program to obtain model which can predict concentration of tropospheric ozone in Bandung city. The relation between ozone precursors concentration and meteorological parameters to ozone concentration was indicated by the correlation value. Positive correlation values, the increase in ozone concentration, was infl uenced by variables of temperature and solar radiation. A negative correlation value, the decline in ozone concentration, is affected by humidity variables. Modelling results indicate that the use of temperature and concentrations NO2 as data input appeared to be a dominant contributors rather than the other input parameters.