THE SQUARE ROOT ENSEMBLE KALMAN FILTER TO ESTIMATE THE CONCENTRATION OF AIR POLLUTION

Kalman filter is an algorithm to estimate the state variable of dynamical stochastic system. The advantages of the Kalman filter are ensemble Kalman filter (EnKF), Square root Ensemble Kalman Filter (SQRT-EnKF) In the ensemble Kalman filter, we need more computational time than the Kalman filter. The square root ensemble Kalman filter is proposed to keep the computational stability and the reduced rank ensemble Kalman filter is proposed to keep the computational stability and reduce the computational time of square root ensemble Kalman filter. We have applied Kalman filter to estimate the concentration of air pollution in the city. Some area in city is divided in n×n grids, some data are taken to estimate for n position. Here we applied the SQRT-EnKF and EnKF algorithm to estimate the concentration of air pollution. We compare the accuracy and computational time between those algorithms.