 The statistical framework of data assimilation allows for the use of new remote sensing data to update forest databases by combining existing data with each new dataset, utilizing all available data in proportion to their quality to improve prediction. Extended Kalman filtering was used in this study on 137 sample plots over a period of four years at a test site in southern Sweden, resulting in predictions closer to the reference value than predictions based on single time points for variables such as lawy's mean height, basal area, and stem volume. The median reduction in root mean square error was 0.4 m, 0.9 m2 per hectare, and 15.3 m3 per hectare, 2%, 3%, and 6%, respectively. This article was authored by Niels Lindgren, Henrik J. Persson, Mathias Nystrom, and others.