An ensemble platform on generalized source tracking of urban air pollutants and an emission source assimilation with application of satellite remote sensing has been proposed. The ensemble platform of generalized urban air pollutant source tracking is built on principal factor analysis of aerosol emissions, aerosol variation spectrum analysis, feature analysis on aerosol changes with phase within urban canopy layer, wind vectors correlating with remote pollutant transport, back forward trajectories and boundary layers affecting the foot mark methods. Also, in order to reduce the emission source uncertainties, which is a key difficulty in air quality modeling, the Newton approximation nudging technology has been applied in the air pollutant modeling based on the seasonal or monthly SO2 and NO2 emissions, and the remote sensing data of air pollutants have been applied on the data assimilation in air pollutant modeling. The integration approach with remote sensing and ground based observations could be used in the research on urban air pollutant tracking, urban haze and heat island. The comprehensive analysis of land water air cycles is attempted as an exploratory study on water pollutant source tracking.