Spatial and temporal deep learning methods for deriving land-use following deforestation: A pan-tropical case study using Landsat time series

Assessing land-use following deforestation is vital for reducing emissions from deforestation and forest degradation. In this paper, for the first time, we assess the potential of spatial, temporal and spatio-temporal deep learning methods for large-scale classification of land-use following tropica...

Descripción completa

Detalles Bibliográficos
Autores principales: Masolele, R.N., Sy, Veronique de, Herold, M., Gonzalez, D.M., Verbesselt, Jan, Gieseke, F., Mullissa, A.G., Martius, C.
Formato: Journal Article
Lenguaje:Inglés
Publicado: Elsevier 2021
Materias:
Acceso en línea:https://hdl.handle.net/10568/115561
_version_ 1855524684280365056
author Masolele, R.N.
Sy, Veronique de
Herold, M.
Gonzalez, D.M.
Verbesselt, Jan
Gieseke, F.
Mullissa, A.G.
Martius, C.
author_browse Gieseke, F.
Gonzalez, D.M.
Herold, M.
Martius, C.
Masolele, R.N.
Mullissa, A.G.
Sy, Veronique de
Verbesselt, Jan
author_facet Masolele, R.N.
Sy, Veronique de
Herold, M.
Gonzalez, D.M.
Verbesselt, Jan
Gieseke, F.
Mullissa, A.G.
Martius, C.
author_sort Masolele, R.N.
collection Repository of Agricultural Research Outputs (CGSpace)
description Assessing land-use following deforestation is vital for reducing emissions from deforestation and forest degradation. In this paper, for the first time, we assess the potential of spatial, temporal and spatio-temporal deep learning methods for large-scale classification of land-use following tropical deforestation using dense satellite time series over six years on the pan-tropical scale (incl. Latin America, Africa, and Asia). Based on an extensive reference database of six forest to land-use conversion types, we find that the spatio-temporal models achieved a substantially higher F1-score accuracies than models that account only for spatial or temporal patterns. Although all models performed better when the scope of the problem was limited to a single continent, the spatial models were more competitive than the temporal ones in this setting. These results suggest that the spatial patterns of land-use within a continent share more commonalities than the temporal patterns and the spatial patterns across continents. This work explores the feasibility of extending and complementing previous efforts for characterizing follow-up land-use after deforestation at a small-scale via human visual interpretation of high resolution RGB imagery. It supports the usage of fast and automated large-scale land-use classification and showcases the value of deep learning methods combined with spatio-temporal satellite data to effectively address the complex tasks of identifying land-use following deforestation in a scalable and cost effective manner.
format Journal Article
id CGSpace115561
institution CGIAR Consortium
language Inglés
publishDate 2021
publishDateRange 2021
publishDateSort 2021
publisher Elsevier
publisherStr Elsevier
record_format dspace
spelling CGSpace1155612024-01-17T12:58:34Z Spatial and temporal deep learning methods for deriving land-use following deforestation: A pan-tropical case study using Landsat time series Masolele, R.N. Sy, Veronique de Herold, M. Gonzalez, D.M. Verbesselt, Jan Gieseke, F. Mullissa, A.G. Martius, C. deforestation land use satellite imagery Assessing land-use following deforestation is vital for reducing emissions from deforestation and forest degradation. In this paper, for the first time, we assess the potential of spatial, temporal and spatio-temporal deep learning methods for large-scale classification of land-use following tropical deforestation using dense satellite time series over six years on the pan-tropical scale (incl. Latin America, Africa, and Asia). Based on an extensive reference database of six forest to land-use conversion types, we find that the spatio-temporal models achieved a substantially higher F1-score accuracies than models that account only for spatial or temporal patterns. Although all models performed better when the scope of the problem was limited to a single continent, the spatial models were more competitive than the temporal ones in this setting. These results suggest that the spatial patterns of land-use within a continent share more commonalities than the temporal patterns and the spatial patterns across continents. This work explores the feasibility of extending and complementing previous efforts for characterizing follow-up land-use after deforestation at a small-scale via human visual interpretation of high resolution RGB imagery. It supports the usage of fast and automated large-scale land-use classification and showcases the value of deep learning methods combined with spatio-temporal satellite data to effectively address the complex tasks of identifying land-use following deforestation in a scalable and cost effective manner. 2021-10 2021-10-21T02:35:53Z 2021-10-21T02:35:53Z Journal Article https://hdl.handle.net/10568/115561 en Open Access Elsevier Masolele, R.N., De Sy, V., Herold, M., Gonzalez, D.M., Verbesselt, J., Gieseke, F., Mullissa, A.G. and Martius, C., 2021. Spatial and temporal deep learning methods for deriving land-use following deforestation: A pan-tropical case study using Landsat time series. Remote Sensing of Environment, 264, 112600. https://doi.org/10.1016/j.rse.2021.112600
spellingShingle deforestation
land use
satellite imagery
Masolele, R.N.
Sy, Veronique de
Herold, M.
Gonzalez, D.M.
Verbesselt, Jan
Gieseke, F.
Mullissa, A.G.
Martius, C.
Spatial and temporal deep learning methods for deriving land-use following deforestation: A pan-tropical case study using Landsat time series
title Spatial and temporal deep learning methods for deriving land-use following deforestation: A pan-tropical case study using Landsat time series
title_full Spatial and temporal deep learning methods for deriving land-use following deforestation: A pan-tropical case study using Landsat time series
title_fullStr Spatial and temporal deep learning methods for deriving land-use following deforestation: A pan-tropical case study using Landsat time series
title_full_unstemmed Spatial and temporal deep learning methods for deriving land-use following deforestation: A pan-tropical case study using Landsat time series
title_short Spatial and temporal deep learning methods for deriving land-use following deforestation: A pan-tropical case study using Landsat time series
title_sort spatial and temporal deep learning methods for deriving land use following deforestation a pan tropical case study using landsat time series
topic deforestation
land use
satellite imagery
url https://hdl.handle.net/10568/115561
work_keys_str_mv AT masolelern spatialandtemporaldeeplearningmethodsforderivinglandusefollowingdeforestationapantropicalcasestudyusinglandsattimeseries
AT syveroniquede spatialandtemporaldeeplearningmethodsforderivinglandusefollowingdeforestationapantropicalcasestudyusinglandsattimeseries
AT heroldm spatialandtemporaldeeplearningmethodsforderivinglandusefollowingdeforestationapantropicalcasestudyusinglandsattimeseries
AT gonzalezdm spatialandtemporaldeeplearningmethodsforderivinglandusefollowingdeforestationapantropicalcasestudyusinglandsattimeseries
AT verbesseltjan spatialandtemporaldeeplearningmethodsforderivinglandusefollowingdeforestationapantropicalcasestudyusinglandsattimeseries
AT giesekef spatialandtemporaldeeplearningmethodsforderivinglandusefollowingdeforestationapantropicalcasestudyusinglandsattimeseries
AT mullissaag spatialandtemporaldeeplearningmethodsforderivinglandusefollowingdeforestationapantropicalcasestudyusinglandsattimeseries
AT martiusc spatialandtemporaldeeplearningmethodsforderivinglandusefollowingdeforestationapantropicalcasestudyusinglandsattimeseries