AN APPLICATION OF KALMAN FILTERING FOR MONITORING FOREST GROWTH AIDED BY SATELLITE IMAGE TIME SERIES

AN APPLICATION OF KALMAN FILTERING FOR MONITORING FOREST GROWTH AIDED BY SATELLITE IMAGE TIME SERIES. Joyce, S. World Scientific Publishing Co. Pte. Ltd., 2002.
doi abstract bibtex

Abstract This paper presents a framework for updating estimates of forest parameters based on historical inventory data and growth models using a time sequence of satellite remote sensing measurements. If a forest growth model can be constructed in linear state equation form, with a vector of forest parameters representing the current state estimate, then it is shown how available remote sensing measurements can be used to improve the current state estimate and reduce its variance using Kalman filtering. Differences in the conditions of image acquisition and image quality are handled implicitly by constructing acquisition-specific models relating spectral measurements to forest parameters. The technique is applied and evaluated on a selection of permanent sample plots from the Swedish National Forest Inventory and a time sequence of 7 Landsat TM scenes over an 11-year period. When the initial forest state is known with some certainty, then the remote sensing measurements give only modest improvement over growth modeling alone, except perhaps when unexpected disturbances occur. However, it is straightforward to apply the same technique when the initial state is totally unknown or initial measurements are poor. In this case, forest parameter estimation and monitoring with remote sensing time series using Kalman filtering may be a useful complement to field inventory activities.

@book{RN201,
   author = {Joyce, Steve},
   title = {AN APPLICATION OF KALMAN FILTERING FOR MONITORING FOREST GROWTH AIDED BY SATELLITE IMAGE TIME SERIES},
   publisher = {World Scientific Publishing Co. Pte. Ltd.},
   abstract = {Abstract This paper presents a framework for updating estimates of forest parameters based on historical inventory data and growth models using a time sequence of satellite remote sensing measurements. If a forest growth model can be constructed in linear state equation form, with a vector of forest parameters representing the current state estimate, then it is shown how available remote sensing measurements can be used to improve the current state estimate and reduce its variance using Kalman filtering. Differences in the conditions of image acquisition and image quality are handled implicitly by constructing acquisition-specific models relating spectral measurements to forest parameters. The technique is applied and evaluated on a selection of permanent sample plots from the Swedish National Forest Inventory and a time sequence of 7 Landsat TM scenes over an 11-year period. When the initial forest state is known with some certainty, then the remote sensing measurements give only modest improvement over growth modeling alone, except perhaps when unexpected disturbances occur. However, it is straightforward to apply the same technique when the initial state is totally unknown or initial measurements are poor. In this case, forest parameter estimation and monitoring with remote sensing time series using Kalman filtering may be a useful complement to field inventory activities.},
   keywords = {Land-Cover Dynamics},
   pages = {371-378},
   DOI = {10.1142/9789812777249_0042},
   year = {2002},
   type = {Book}
}

Downloads: 0

{"_id":"Rexy8CDmZRAzfb8Y5","bibbaseid":"joyce-anapplicationofkalmanfilteringformonitoringforestgrowthaidedbysatelliteimagetimeseries-2002","author_short":["Joyce, S."],"bibdata":{"bibtype":"book","type":"Book","author":[{"propositions":[],"lastnames":["Joyce"],"firstnames":["Steve"],"suffixes":[]}],"title":"AN APPLICATION OF KALMAN FILTERING FOR MONITORING FOREST GROWTH AIDED BY SATELLITE IMAGE TIME SERIES","publisher":"World Scientific Publishing Co. Pte. Ltd.","abstract":"Abstract This paper presents a framework for updating estimates of forest parameters based on historical inventory data and growth models using a time sequence of satellite remote sensing measurements. If a forest growth model can be constructed in linear state equation form, with a vector of forest parameters representing the current state estimate, then it is shown how available remote sensing measurements can be used to improve the current state estimate and reduce its variance using Kalman filtering. Differences in the conditions of image acquisition and image quality are handled implicitly by constructing acquisition-specific models relating spectral measurements to forest parameters. The technique is applied and evaluated on a selection of permanent sample plots from the Swedish National Forest Inventory and a time sequence of 7 Landsat TM scenes over an 11-year period. When the initial forest state is known with some certainty, then the remote sensing measurements give only modest improvement over growth modeling alone, except perhaps when unexpected disturbances occur. However, it is straightforward to apply the same technique when the initial state is totally unknown or initial measurements are poor. In this case, forest parameter estimation and monitoring with remote sensing time series using Kalman filtering may be a useful complement to field inventory activities.","keywords":"Land-Cover Dynamics","pages":"371-378","doi":"10.1142/9789812777249_0042","year":"2002","bibtex":"@book{RN201,\r\n author = {Joyce, Steve},\r\n title = {AN APPLICATION OF KALMAN FILTERING FOR MONITORING FOREST GROWTH AIDED BY SATELLITE IMAGE TIME SERIES},\r\n publisher = {World Scientific Publishing Co. Pte. Ltd.},\r\n abstract = {Abstract This paper presents a framework for updating estimates of forest parameters based on historical inventory data and growth models using a time sequence of satellite remote sensing measurements. If a forest growth model can be constructed in linear state equation form, with a vector of forest parameters representing the current state estimate, then it is shown how available remote sensing measurements can be used to improve the current state estimate and reduce its variance using Kalman filtering. Differences in the conditions of image acquisition and image quality are handled implicitly by constructing acquisition-specific models relating spectral measurements to forest parameters. The technique is applied and evaluated on a selection of permanent sample plots from the Swedish National Forest Inventory and a time sequence of 7 Landsat TM scenes over an 11-year period. When the initial forest state is known with some certainty, then the remote sensing measurements give only modest improvement over growth modeling alone, except perhaps when unexpected disturbances occur. However, it is straightforward to apply the same technique when the initial state is totally unknown or initial measurements are poor. In this case, forest parameter estimation and monitoring with remote sensing time series using Kalman filtering may be a useful complement to field inventory activities.},\r\n keywords = {Land-Cover Dynamics},\r\n pages = {371-378},\r\n DOI = {10.1142/9789812777249_0042},\r\n year = {2002},\r\n type = {Book}\r\n}\r\n\r\n","author_short":["Joyce, S."],"key":"RN201","id":"RN201","bibbaseid":"joyce-anapplicationofkalmanfilteringformonitoringforestgrowthaidedbysatelliteimagetimeseries-2002","role":"author","urls":{},"keyword":["Land-Cover Dynamics"],"metadata":{"authorlinks":{}}},"bibtype":"book","biburl":"https://www.slu.se/globalassets/ew/org/centrb/rt/dokument/publikationslistor/rikskogstaxeringsprojektet_bibtex_ny.txt","dataSources":["Cac5nog9ND5ndLYhY","xTmt4jq9swAwpBHJB","LXbacBrgTRDPkh9C2","dLPsL5XH9N5Pjush7"],"keywords":["land-cover dynamics"],"search_terms":["application","kalman","filtering","monitoring","forest","growth","aided","satellite","image","time","series","joyce"],"title":"AN APPLICATION OF KALMAN FILTERING FOR MONITORING FOREST GROWTH AIDED BY SATELLITE IMAGE TIME SERIES","year":2002}