Estimation of atmospheric CO2concentration using Kalman filtering

Abstract
Using a dynamic state model for the observed upward trend and sinusoidal variation, a Kalman filter is constructed to estimate the atmospheric CO2 concentration, The process noise is assumed to be white with an unknown covariance, so an adaptive scheme is used to estimate the steady-state Kalman gain matrix. Several tests for optimality are performed on the adaptive filter. Measured data are then filtered using the Kalman algorithm. The filtering results are shown to reduce the variability of the airborne fraction of fossil-fuel-produced atmospheric CO2.

This publication has 15 references indexed in Scilit: