Using high-resolution imagery and deep learning to classify land-use following deforestation: a case study in Ethiopia
National-scale assessments of post-deforestation land-use are crucial for decreasing deforestation and forest degradation-related emissions. In this research, we assess the potential of different satellite data modalities (single-date, multi-date, multi-resolution, and an ensemble of multi-sensor im...
| Autores principales: | , , , , , , , , , |
|---|---|
| Formato: | Journal Article |
| Lenguaje: | Inglés |
| Publicado: |
Informa UK Limited
2022
|
| Materias: | |
| Acceso en línea: | https://hdl.handle.net/10568/125031 |
| _version_ | 1855519815921303552 |
|---|---|
| author | Masolele, Robert N. Sy, Veronique de Marcos, Diego Verbesselt, Jan Gieseke, Fabian Mulatu, Kalkidan Ayele Moges, Yitebitu Sebrala, Heiru Martius, Christopher Herold, Martin |
| author_browse | Gieseke, Fabian Herold, Martin Marcos, Diego Martius, Christopher Masolele, Robert N. Moges, Yitebitu Mulatu, Kalkidan Ayele Sebrala, Heiru Sy, Veronique de Verbesselt, Jan |
| author_facet | Masolele, Robert N. Sy, Veronique de Marcos, Diego Verbesselt, Jan Gieseke, Fabian Mulatu, Kalkidan Ayele Moges, Yitebitu Sebrala, Heiru Martius, Christopher Herold, Martin |
| author_sort | Masolele, Robert N. |
| collection | Repository of Agricultural Research Outputs (CGSpace) |
| description | National-scale assessments of post-deforestation land-use are crucial for decreasing deforestation and forest degradation-related emissions. In this research, we assess the potential of different satellite data modalities (single-date, multi-date, multi-resolution, and an ensemble of multi-sensor images) for classifying land-use following deforestation in Ethiopia using the U-Net deep neural network architecture enhanced with attention. We performed the analysis on satellite image data retrieved across Ethiopia from freely available Landsat-8, Sentinel-2 and Planet-NICFI satellite data. The experiments aimed at an analysis of (a) single-date images from individual sensors to account for the differences in spatial resolution between image sensors in detecting land-uses, (b) ensembles of multiple images from different sensors (Planet-NICFI/Sentinel-2/Landsat-8) with different spatial resolutions, (c) the use of multi-date data to account for the contribution of temporal information in detecting land-uses, and, finally, (d) the identification of regional differences in terms of land-use following deforestation in Ethiopia. We hypothesize that choosing the right satellite imagery (sensor) type is crucial for the task. Based on a comprehensive visually interpreted reference dataset of 11 types of post-deforestation land-uses, we find that either detailed spatial patterns (single-date Planet-NICFI) or detailed temporal patterns (multi-date Sentinel-2, Landsat-8) are required for identifying land-use following deforestation, while medium-resolution single-date imagery is not sufficient to achieve high classification accuracy. We also find that adding soft-attention to the standard U-Net improved the classification accuracy, especially for small-scale land-uses. The models and products presented in this work can be used as a powerful data resource for governmental and forest monitoring agencies to design and monitor deforestation mitigation measures and data-driven land-use policy. |
| format | Journal Article |
| id | CGSpace125031 |
| institution | CGIAR Consortium |
| language | Inglés |
| publishDate | 2022 |
| publishDateRange | 2022 |
| publishDateSort | 2022 |
| publisher | Informa UK Limited |
| publisherStr | Informa UK Limited |
| record_format | dspace |
| spelling | CGSpace1250312025-10-26T12:55:01Z Using high-resolution imagery and deep learning to classify land-use following deforestation: a case study in Ethiopia Masolele, Robert N. Sy, Veronique de Marcos, Diego Verbesselt, Jan Gieseke, Fabian Mulatu, Kalkidan Ayele Moges, Yitebitu Sebrala, Heiru Martius, Christopher Herold, Martin deforestation forestry geographical information systems remote sensing climate change satellite imagery National-scale assessments of post-deforestation land-use are crucial for decreasing deforestation and forest degradation-related emissions. In this research, we assess the potential of different satellite data modalities (single-date, multi-date, multi-resolution, and an ensemble of multi-sensor images) for classifying land-use following deforestation in Ethiopia using the U-Net deep neural network architecture enhanced with attention. We performed the analysis on satellite image data retrieved across Ethiopia from freely available Landsat-8, Sentinel-2 and Planet-NICFI satellite data. The experiments aimed at an analysis of (a) single-date images from individual sensors to account for the differences in spatial resolution between image sensors in detecting land-uses, (b) ensembles of multiple images from different sensors (Planet-NICFI/Sentinel-2/Landsat-8) with different spatial resolutions, (c) the use of multi-date data to account for the contribution of temporal information in detecting land-uses, and, finally, (d) the identification of regional differences in terms of land-use following deforestation in Ethiopia. We hypothesize that choosing the right satellite imagery (sensor) type is crucial for the task. Based on a comprehensive visually interpreted reference dataset of 11 types of post-deforestation land-uses, we find that either detailed spatial patterns (single-date Planet-NICFI) or detailed temporal patterns (multi-date Sentinel-2, Landsat-8) are required for identifying land-use following deforestation, while medium-resolution single-date imagery is not sufficient to achieve high classification accuracy. We also find that adding soft-attention to the standard U-Net improved the classification accuracy, especially for small-scale land-uses. The models and products presented in this work can be used as a powerful data resource for governmental and forest monitoring agencies to design and monitor deforestation mitigation measures and data-driven land-use policy. 2022-12-31 2022-10-13T14:25:33Z 2022-10-13T14:25:33Z Journal Article https://hdl.handle.net/10568/125031 en Open Access Informa UK Limited Masolele, R. N., De Sy, V., Marcos, D., Verbesselt, J., Gieseke, F., Mulatu, K. A., Moges, Y., Sebrala, H., Martius, C., & Herold, M. (2022). Using high-resolution imagery and deep learning to classify land-use following deforestation: a case study in Ethiopia. In GIScience & Remote Sensing (Vol. 59, Issue 1, pp. 1446–1472). Informa UK Limited. https://doi.org/10.1080/15481603.2022.2115619 |
| spellingShingle | deforestation forestry geographical information systems remote sensing climate change satellite imagery Masolele, Robert N. Sy, Veronique de Marcos, Diego Verbesselt, Jan Gieseke, Fabian Mulatu, Kalkidan Ayele Moges, Yitebitu Sebrala, Heiru Martius, Christopher Herold, Martin Using high-resolution imagery and deep learning to classify land-use following deforestation: a case study in Ethiopia |
| title | Using high-resolution imagery and deep learning to classify land-use following deforestation: a case study in Ethiopia |
| title_full | Using high-resolution imagery and deep learning to classify land-use following deforestation: a case study in Ethiopia |
| title_fullStr | Using high-resolution imagery and deep learning to classify land-use following deforestation: a case study in Ethiopia |
| title_full_unstemmed | Using high-resolution imagery and deep learning to classify land-use following deforestation: a case study in Ethiopia |
| title_short | Using high-resolution imagery and deep learning to classify land-use following deforestation: a case study in Ethiopia |
| title_sort | using high resolution imagery and deep learning to classify land use following deforestation a case study in ethiopia |
| topic | deforestation forestry geographical information systems remote sensing climate change satellite imagery |
| url | https://hdl.handle.net/10568/125031 |
| work_keys_str_mv | AT masolelerobertn usinghighresolutionimageryanddeeplearningtoclassifylandusefollowingdeforestationacasestudyinethiopia AT syveroniquede usinghighresolutionimageryanddeeplearningtoclassifylandusefollowingdeforestationacasestudyinethiopia AT marcosdiego usinghighresolutionimageryanddeeplearningtoclassifylandusefollowingdeforestationacasestudyinethiopia AT verbesseltjan usinghighresolutionimageryanddeeplearningtoclassifylandusefollowingdeforestationacasestudyinethiopia AT giesekefabian usinghighresolutionimageryanddeeplearningtoclassifylandusefollowingdeforestationacasestudyinethiopia AT mulatukalkidanayele usinghighresolutionimageryanddeeplearningtoclassifylandusefollowingdeforestationacasestudyinethiopia AT mogesyitebitu usinghighresolutionimageryanddeeplearningtoclassifylandusefollowingdeforestationacasestudyinethiopia AT sebralaheiru usinghighresolutionimageryanddeeplearningtoclassifylandusefollowingdeforestationacasestudyinethiopia AT martiuschristopher usinghighresolutionimageryanddeeplearningtoclassifylandusefollowingdeforestationacasestudyinethiopia AT heroldmartin usinghighresolutionimageryanddeeplearningtoclassifylandusefollowingdeforestationacasestudyinethiopia |