Guillaume Cyril Henri Goodwin
Research Subjects
Guillaume Cyril Henri Goodwin has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Andrea D'Alpaos
Department of Geosciences University of Padova Padova Italy
8 shared publications - Marco Marani
Department of Civil, Architectural, and Environmental Engineering University of Padova Padova Italy
8 shared publications - Sonia Silvestri
University of Bologna (Bologna, Italy)
8 shared publications Luca Carniello
DICEA, Universita di Padova, via Marzolo, 9, Padova, Italy
6 shared publications- Luca Carniello
Department of Civil, Environmental, and Architectural Engineering University of Padova Padova Italy
2 shared publications - Alice Lefebvre
MARUM – Center for Marine Environmental Sciences, University of Bremen, Bremen, Germany
1 shared publication
Latest Publications
- Open Access
- 28 May 2024
- Published by Penn State University Libraries in Geomorphica
- Vol. 1 (1)
Geomorphica: The most accessible journal for the geomorphology community
Geomorphica is a peer-reviewed diamond open access journal for the geosciences, which allows for free access and publication without any subscription fees or article processing charges, thereby enhancing inclusivity and removing obstacles in scientific communication. Geomorphica is both community-led and community-driven, dedicated to sharing new insights, data and promoting advancements in the field of geomorphology. It is hosted by Penn State University Libraries and receives support from the International Association of Geomorphologists. In the last two years, a team of over 30 volunteers has contributed to various aspects of journal development and management, including administration, editing, reviewing, typesetting, and visual branding. Since its launch in June 2023, Geomorphica invites manuscript submissions on topics related to landscapes and landforms, near-surface processes of Earth and other planets, as well as the mechanisms, dynamics, and timescales associated with these processes. We encourage experts in these fields to join our reviewer database or reach out to the journal if they wish to volunteer. To mitigate bias, we have implemented a triple-anonymous review process, ensuring that authors, reviewers, and the associate editor making recommendations remain unknown to one another. The reports from reviewers and the responses from authors are made publicly accessible alongside the published articles, with reviewers having the option to maintain their anonymity. We invite feedback from the wider community to enhance Geomorphica continually and look forward to collaborative engagement with this initiative.
- 28 May 2024
- Published by Penn State University Libraries in Geomorphica
- Vol. 1 (1)
Open AccessGeomorphica: The most accessible journal for the geomorphology community
Geomorphica is a peer-reviewed diamond open access journal for the geosciences, which allows for free access and publication without any subscription fees or article processing charges, thereby enhancing inclusivity and removing obstacles in scientific communication. Geomorphica is both community-led and community-driven, dedicated to sharing new insights, data and promoting advancements in the field of geomorphology. It is hosted by Penn State University Libraries and receives support from the International Association of Geomorphologists. In the last two years, a team of over 30 volunteers has contributed to various aspects of journal development and management, including administration, editing, reviewing, typesetting, and visual branding. Since its launch in June 2023, Geomorphica invites manuscript submissions on topics related to landscapes and landforms, near-surface processes of Earth and other planets, as well as the mechanisms, dynamics, and timescales associated with these processes. We encourage experts in these fields to join our reviewer database or reach out to the journal if they wish to volunteer. To mitigate bias, we have implemented a triple-anonymous review process, ensuring that authors, reviewers, and the associate editor making recommendations remain unknown to one another. The reports from reviewers and the responses from authors are made publicly accessible alongside the published articles, with reviewers having the option to maintain their anonymity. We invite feedback from the wider community to enhance Geomorphica continually and look forward to collaborative engagement with this initiative.
- Open Access
- 20 November 2023
- Published by Copernicus GmbH in Biogeosciences
- Vol. 20 (22) , 4551-4576
Toward coherent space–time mapping of seagrass cover from satellite data: an example of a Mediterranean lagoon
Seagrass meadows are a highly productive and economically important shallow coastal habitat. Their sensitivity to natural and anthropogenic disturbances, combined with their importance for local biodiversity, carbon stocks, and sediment dynamics, motivate a frequent monitoring of their distribution. However, generating time series of seagrass cover from field observations is costly, and mapping methods based on remote sensing require restrictive conditions on seabed visibility, limiting the frequency of observations. In this contribution, we examine the effect of accounting for environmental factors, such as the bathymetry and median grain size (D50) of the substrate as well as the coordinates of known seagrass patches, on the performance of a random forest (RF) classifier used to determine seagrass cover. Using 148 Landsat images of the Venice Lagoon (Italy) between 1999 and 2020, we trained an RF classifier with only spectral features from Landsat images and seagrass surveys from 2002 and 2017. Then, by adding the features above and applying a time-based correction to predictions, we created multiple RF models with different feature combinations. We tested the quality of the resulting seagrass cover predictions from each model against field surveys, showing that bathymetry, D50, and coordinates of known patches exert an influence that is dependent on the training Landsat image and seagrass survey chosen. In models trained on a survey from 2017, where using only spectral features causes predictions to overestimate seagrass surface area, no significant change in model performance was observed. Conversely, in models trained on a survey from 2002, the addition of the out-of-image features and particularly coordinates of known vegetated patches greatly improves the predictive capacity of the model, while still allowing the detection of seagrass beds absent in the reference field survey. Applying a time-based correction eliminates small temporal variations in predictions, improving predictions that performed well before correction. We conclude that accounting for the coordinates of known seagrass patches, together with applying a time-based correction, has the most potential to produce reliable frequent predictions of seagrass cover. While this case study alone is insufficient to explain how geographic location information influences the classification process, we suggest that it is linked to the inherent spatial auto-correlation of seagrass meadow distribution. In the interest of improving remote-sensing classification and particularly to develop our capacity to map vegetation across time, we identify this phenomenon as warranting further research.
- 20 November 2023
- Published by Copernicus GmbH in Biogeosciences
- Vol. 20 (22) , 4551-4576
Open AccessToward coherent space–time mapping of seagrass cover from satellite data: an example of a Mediterranean lagoon
Seagrass meadows are a highly productive and economically important shallow coastal habitat. Their sensitivity to natural and anthropogenic disturbances, combined with their importance for local biodiversity, carbon stocks, and sediment dynamics, motivate a frequent monitoring of their distribution. However, generating time series of seagrass cover from field observations is costly, and mapping methods based on remote sensing require restrictive conditions on seabed visibility, limiting the frequency of observations. In this contribution, we examine the effect of accounting for environmental factors, such as the bathymetry and median grain size (D50) of the substrate as well as the coordinates of known seagrass patches, on the performance of a random forest (RF) classifier used to determine seagrass cover. Using 148 Landsat images of the Venice Lagoon (Italy) between 1999 and 2020, we trained an RF classifier with only spectral features from Landsat images and seagrass surveys from 2002 and 2017. Then, by adding the features above and applying a time-based correction to predictions, we created multiple RF models with different feature combinations. We tested the quality of the resulting seagrass cover predictions from each model against field surveys, showing that bathymetry, D50, and coordinates of known patches exert an influence that is dependent on the training Landsat image and seagrass survey chosen. In models trained on a survey from 2017, where using only spectral features causes predictions to overestimate seagrass surface area, no significant change in model performance was observed. Conversely, in models trained on a survey from 2002, the addition of the out-of-image features and particularly coordinates of known vegetated patches greatly improves the predictive capacity of the model, while still allowing the detection of seagrass beds absent in the reference field survey. Applying a time-based correction eliminates small temporal variations in predictions, improving predictions that performed well before correction. We conclude that accounting for the coordinates of known seagrass patches, together with applying a time-based correction, has the most potential to produce reliable frequent predictions of seagrass cover. While this case study alone is insufficient to explain how geographic location information influences the classification process, we suggest that it is linked to the inherent spatial auto-correlation of seagrass meadow distribution. In the interest of improving remote-sensing classification and particularly to develop our capacity to map vegetation across time, we identify this phenomenon as warranting further research.
- Open Access
- 25 April 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-32
Reply on RC1
Seagrass meadows are a highly productive and economically important shallow coastal habitat. Their sensitivity to natural and anthropogenic disturbances, combined with their importance for local biodiversity, carbon stocks and sediment dynamics, motivate a frequent monitoring of their distribution. However, generating time-series of seagrass cover from field observations is costly, and mapping methods based on remote sensing require restrictive conditions on seabed visibility, limiting the frequency of observations. In this contribution, we examine the effect of accounting for environmental factors such as the bathymetry and median grain size (D50) of the substrate, as well as the coordinates of known seagrass patches, on the performance of a Random Forest (RF) classifier used to determine seagrass cover. Using 148 Landsat images of the Venice Lagoon (Italy) between 1999 and 2020, we trained a RF classifier with only spectral features from Landsat images and seagrass surveys, respectively from 2002 and 2017. Then, by adding the features above and applying a time-based correction on predictions, we created multiple RF models with different feature combinations. We tested the quality of the resulting seagrass cover predictions from each model against field surveys, showing that bathymetry, D50 and coordinates of known patches exert an influence that is dependant on the training Landsat image and seagrass survey chosen. In models trained on a survey from 2017, where using only spectral features causes predictions to overestimate seagrass surface area, no significant change in model performance was observed. Conversely, in models trained on a survey from 2002, the addition of the out-of-image features and particularly coordinates of known vegetated patches greatly improves the predictive capacity of the model, while still allowing the detection of seagrass beds absent in the reference field survey. Applying a time-based correction eliminates small temporal variations in predictions, improving predictions that performed well before correction. We conclude that accounting for the coordinates of known seagrass patches, together with applying a time-based correction, has the most potential to produce reliable frequent predictions of seagrass cover. While this case study alone is insufficient to explain how geographic location information influences the classification process, we suggest that it is linked to the inherent spatial auto-correlation of seagrass meadow distribution. In the interest of improving remote sensing classification and particularly to develop our capacity to map vegetation across time, we identify this phenomenon as warranting further research.
- 25 April 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-32
Open AccessReply on RC1
Seagrass meadows are a highly productive and economically important shallow coastal habitat. Their sensitivity to natural and anthropogenic disturbances, combined with their importance for local biodiversity, carbon stocks and sediment dynamics, motivate a frequent monitoring of their distribution. However, generating time-series of seagrass cover from field observations is costly, and mapping methods based on remote sensing require restrictive conditions on seabed visibility, limiting the frequency of observations. In this contribution, we examine the effect of accounting for environmental factors such as the bathymetry and median grain size (D50) of the substrate, as well as the coordinates of known seagrass patches, on the performance of a Random Forest (RF) classifier used to determine seagrass cover. Using 148 Landsat images of the Venice Lagoon (Italy) between 1999 and 2020, we trained a RF classifier with only spectral features from Landsat images and seagrass surveys, respectively from 2002 and 2017. Then, by adding the features above and applying a time-based correction on predictions, we created multiple RF models with different feature combinations. We tested the quality of the resulting seagrass cover predictions from each model against field surveys, showing that bathymetry, D50 and coordinates of known patches exert an influence that is dependant on the training Landsat image and seagrass survey chosen. In models trained on a survey from 2017, where using only spectral features causes predictions to overestimate seagrass surface area, no significant change in model performance was observed. Conversely, in models trained on a survey from 2002, the addition of the out-of-image features and particularly coordinates of known vegetated patches greatly improves the predictive capacity of the model, while still allowing the detection of seagrass beds absent in the reference field survey. Applying a time-based correction eliminates small temporal variations in predictions, improving predictions that performed well before correction. We conclude that accounting for the coordinates of known seagrass patches, together with applying a time-based correction, has the most potential to produce reliable frequent predictions of seagrass cover. While this case study alone is insufficient to explain how geographic location information influences the classification process, we suggest that it is linked to the inherent spatial auto-correlation of seagrass meadow distribution. In the interest of improving remote sensing classification and particularly to develop our capacity to map vegetation across time, we identify this phenomenon as warranting further research.
- Open Access
- 25 April 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-32
Reply on RC2
Seagrass meadows are a highly productive and economically important shallow coastal habitat. Their sensitivity to natural and anthropogenic disturbances, combined with their importance for local biodiversity, carbon stocks and sediment dynamics, motivate a frequent monitoring of their distribution. However, generating time-series of seagrass cover from field observations is costly, and mapping methods based on remote sensing require restrictive conditions on seabed visibility, limiting the frequency of observations. In this contribution, we examine the effect of accounting for environmental factors such as the bathymetry and median grain size (D50) of the substrate, as well as the coordinates of known seagrass patches, on the performance of a Random Forest (RF) classifier used to determine seagrass cover. Using 148 Landsat images of the Venice Lagoon (Italy) between 1999 and 2020, we trained a RF classifier with only spectral features from Landsat images and seagrass surveys, respectively from 2002 and 2017. Then, by adding the features above and applying a time-based correction on predictions, we created multiple RF models with different feature combinations. We tested the quality of the resulting seagrass cover predictions from each model against field surveys, showing that bathymetry, D50 and coordinates of known patches exert an influence that is dependant on the training Landsat image and seagrass survey chosen. In models trained on a survey from 2017, where using only spectral features causes predictions to overestimate seagrass surface area, no significant change in model performance was observed. Conversely, in models trained on a survey from 2002, the addition of the out-of-image features and particularly coordinates of known vegetated patches greatly improves the predictive capacity of the model, while still allowing the detection of seagrass beds absent in the reference field survey. Applying a time-based correction eliminates small temporal variations in predictions, improving predictions that performed well before correction. We conclude that accounting for the coordinates of known seagrass patches, together with applying a time-based correction, has the most potential to produce reliable frequent predictions of seagrass cover. While this case study alone is insufficient to explain how geographic location information influences the classification process, we suggest that it is linked to the inherent spatial auto-correlation of seagrass meadow distribution. In the interest of improving remote sensing classification and particularly to develop our capacity to map vegetation across time, we identify this phenomenon as warranting further research.
- 25 April 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-32
Open AccessReply on RC2
Seagrass meadows are a highly productive and economically important shallow coastal habitat. Their sensitivity to natural and anthropogenic disturbances, combined with their importance for local biodiversity, carbon stocks and sediment dynamics, motivate a frequent monitoring of their distribution. However, generating time-series of seagrass cover from field observations is costly, and mapping methods based on remote sensing require restrictive conditions on seabed visibility, limiting the frequency of observations. In this contribution, we examine the effect of accounting for environmental factors such as the bathymetry and median grain size (D50) of the substrate, as well as the coordinates of known seagrass patches, on the performance of a Random Forest (RF) classifier used to determine seagrass cover. Using 148 Landsat images of the Venice Lagoon (Italy) between 1999 and 2020, we trained a RF classifier with only spectral features from Landsat images and seagrass surveys, respectively from 2002 and 2017. Then, by adding the features above and applying a time-based correction on predictions, we created multiple RF models with different feature combinations. We tested the quality of the resulting seagrass cover predictions from each model against field surveys, showing that bathymetry, D50 and coordinates of known patches exert an influence that is dependant on the training Landsat image and seagrass survey chosen. In models trained on a survey from 2017, where using only spectral features causes predictions to overestimate seagrass surface area, no significant change in model performance was observed. Conversely, in models trained on a survey from 2002, the addition of the out-of-image features and particularly coordinates of known vegetated patches greatly improves the predictive capacity of the model, while still allowing the detection of seagrass beds absent in the reference field survey. Applying a time-based correction eliminates small temporal variations in predictions, improving predictions that performed well before correction. We conclude that accounting for the coordinates of known seagrass patches, together with applying a time-based correction, has the most potential to produce reliable frequent predictions of seagrass cover. While this case study alone is insufficient to explain how geographic location information influences the classification process, we suggest that it is linked to the inherent spatial auto-correlation of seagrass meadow distribution. In the interest of improving remote sensing classification and particularly to develop our capacity to map vegetation across time, we identify this phenomenon as warranting further research.
- Open Access
- 21 February 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-32
Comment on egusphere-2022-1501
Seagrass meadows are a highly productive and economically important shallow coastal habitat. Their sensitivity to natural and anthropogenic disturbances, combined with their importance for local biodiversity, carbon stocks and sediment dynamics, motivate a frequent monitoring of their distribution. However, generating time-series of seagrass cover from field observations is costly, and mapping methods based on remote sensing require restrictive conditions on seabed visibility, limiting the frequency of observations. In this contribution, we examine the effect of accounting for environmental factors such as the bathymetry and median grain size (D50) of the substrate, as well as the coordinates of known seagrass patches, on the performance of a Random Forest (RF) classifier used to determine seagrass cover. Using 148 Landsat images of the Venice Lagoon (Italy) between 1999 and 2020, we trained a RF classifier with only spectral features from Landsat images and seagrass surveys, respectively from 2002 and 2017. Then, by adding the features above and applying a time-based correction on predictions, we created multiple RF models with different feature combinations. We tested the quality of the resulting seagrass cover predictions from each model against field surveys, showing that bathymetry, D50 and coordinates of known patches exert an influence that is dependant on the training Landsat image and seagrass survey chosen. In models trained on a survey from 2017, where using only spectral features causes predictions to overestimate seagrass surface area, no significant change in model performance was observed. Conversely, in models trained on a survey from 2002, the addition of the out-of-image features and particularly coordinates of known vegetated patches greatly improves the predictive capacity of the model, while still allowing the detection of seagrass beds absent in the reference field survey. Applying a time-based correction eliminates small temporal variations in predictions, improving predictions that performed well before correction. We conclude that accounting for the coordinates of known seagrass patches, together with applying a time-based correction, has the most potential to produce reliable frequent predictions of seagrass cover. While this case study alone is insufficient to explain how geographic location information influences the classification process, we suggest that it is linked to the inherent spatial auto-correlation of seagrass meadow distribution. In the interest of improving remote sensing classification and particularly to develop our capacity to map vegetation across time, we identify this phenomenon as warranting further research.
- 21 February 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-32
Open AccessComment on egusphere-2022-1501
Seagrass meadows are a highly productive and economically important shallow coastal habitat. Their sensitivity to natural and anthropogenic disturbances, combined with their importance for local biodiversity, carbon stocks and sediment dynamics, motivate a frequent monitoring of their distribution. However, generating time-series of seagrass cover from field observations is costly, and mapping methods based on remote sensing require restrictive conditions on seabed visibility, limiting the frequency of observations. In this contribution, we examine the effect of accounting for environmental factors such as the bathymetry and median grain size (D50) of the substrate, as well as the coordinates of known seagrass patches, on the performance of a Random Forest (RF) classifier used to determine seagrass cover. Using 148 Landsat images of the Venice Lagoon (Italy) between 1999 and 2020, we trained a RF classifier with only spectral features from Landsat images and seagrass surveys, respectively from 2002 and 2017. Then, by adding the features above and applying a time-based correction on predictions, we created multiple RF models with different feature combinations. We tested the quality of the resulting seagrass cover predictions from each model against field surveys, showing that bathymetry, D50 and coordinates of known patches exert an influence that is dependant on the training Landsat image and seagrass survey chosen. In models trained on a survey from 2017, where using only spectral features causes predictions to overestimate seagrass surface area, no significant change in model performance was observed. Conversely, in models trained on a survey from 2002, the addition of the out-of-image features and particularly coordinates of known vegetated patches greatly improves the predictive capacity of the model, while still allowing the detection of seagrass beds absent in the reference field survey. Applying a time-based correction eliminates small temporal variations in predictions, improving predictions that performed well before correction. We conclude that accounting for the coordinates of known seagrass patches, together with applying a time-based correction, has the most potential to produce reliable frequent predictions of seagrass cover. While this case study alone is insufficient to explain how geographic location information influences the classification process, we suggest that it is linked to the inherent spatial auto-correlation of seagrass meadow distribution. In the interest of improving remote sensing classification and particularly to develop our capacity to map vegetation across time, we identify this phenomenon as warranting further research.
- Open Access
- 15 February 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-32
Comment on egusphere-2022-1501
Seagrass meadows are a highly productive and economically important shallow coastal habitat. Their sensitivity to natural and anthropogenic disturbances, combined with their importance for local biodiversity, carbon stocks and sediment dynamics, motivate a frequent monitoring of their distribution. However, generating time-series of seagrass cover from field observations is costly, and mapping methods based on remote sensing require restrictive conditions on seabed visibility, limiting the frequency of observations. In this contribution, we examine the effect of accounting for environmental factors such as the bathymetry and median grain size (D50) of the substrate, as well as the coordinates of known seagrass patches, on the performance of a Random Forest (RF) classifier used to determine seagrass cover. Using 148 Landsat images of the Venice Lagoon (Italy) between 1999 and 2020, we trained a RF classifier with only spectral features from Landsat images and seagrass surveys, respectively from 2002 and 2017. Then, by adding the features above and applying a time-based correction on predictions, we created multiple RF models with different feature combinations. We tested the quality of the resulting seagrass cover predictions from each model against field surveys, showing that bathymetry, D50 and coordinates of known patches exert an influence that is dependant on the training Landsat image and seagrass survey chosen. In models trained on a survey from 2017, where using only spectral features causes predictions to overestimate seagrass surface area, no significant change in model performance was observed. Conversely, in models trained on a survey from 2002, the addition of the out-of-image features and particularly coordinates of known vegetated patches greatly improves the predictive capacity of the model, while still allowing the detection of seagrass beds absent in the reference field survey. Applying a time-based correction eliminates small temporal variations in predictions, improving predictions that performed well before correction. We conclude that accounting for the coordinates of known seagrass patches, together with applying a time-based correction, has the most potential to produce reliable frequent predictions of seagrass cover. While this case study alone is insufficient to explain how geographic location information influences the classification process, we suggest that it is linked to the inherent spatial auto-correlation of seagrass meadow distribution. In the interest of improving remote sensing classification and particularly to develop our capacity to map vegetation across time, we identify this phenomenon as warranting further research.
- 15 February 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-32
Open AccessComment on egusphere-2022-1501
Seagrass meadows are a highly productive and economically important shallow coastal habitat. Their sensitivity to natural and anthropogenic disturbances, combined with their importance for local biodiversity, carbon stocks and sediment dynamics, motivate a frequent monitoring of their distribution. However, generating time-series of seagrass cover from field observations is costly, and mapping methods based on remote sensing require restrictive conditions on seabed visibility, limiting the frequency of observations. In this contribution, we examine the effect of accounting for environmental factors such as the bathymetry and median grain size (D50) of the substrate, as well as the coordinates of known seagrass patches, on the performance of a Random Forest (RF) classifier used to determine seagrass cover. Using 148 Landsat images of the Venice Lagoon (Italy) between 1999 and 2020, we trained a RF classifier with only spectral features from Landsat images and seagrass surveys, respectively from 2002 and 2017. Then, by adding the features above and applying a time-based correction on predictions, we created multiple RF models with different feature combinations. We tested the quality of the resulting seagrass cover predictions from each model against field surveys, showing that bathymetry, D50 and coordinates of known patches exert an influence that is dependant on the training Landsat image and seagrass survey chosen. In models trained on a survey from 2017, where using only spectral features causes predictions to overestimate seagrass surface area, no significant change in model performance was observed. Conversely, in models trained on a survey from 2002, the addition of the out-of-image features and particularly coordinates of known vegetated patches greatly improves the predictive capacity of the model, while still allowing the detection of seagrass beds absent in the reference field survey. Applying a time-based correction eliminates small temporal variations in predictions, improving predictions that performed well before correction. We conclude that accounting for the coordinates of known seagrass patches, together with applying a time-based correction, has the most potential to produce reliable frequent predictions of seagrass cover. While this case study alone is insufficient to explain how geographic location information influences the classification process, we suggest that it is linked to the inherent spatial auto-correlation of seagrass meadow distribution. In the interest of improving remote sensing classification and particularly to develop our capacity to map vegetation across time, we identify this phenomenon as warranting further research.
- Open Access
- 24 January 2023
- Published by Copernicus GmbH in EGUsphere
- Vol. 2023, 1-32
Toward coherent space-time mapping of seagrass cover from satellite data: example of a Mediterranean lagoon
Seagrass meadows are a highly productive and economically important shallow coastal habitat. Their sensitivity to natural and anthropogenic disturbances, combined with their importance for local biodiversity, carbon stocks and sediment dynamics, motivate a frequent monitoring of their distribution. However, generating time-series of seagrass cover from field observations is costly, and mapping methods based on remote sensing require restrictive conditions on seabed visibility, limiting the frequency of observations. In this contribution, we examine the effect of accounting for environmental factors such as the bathymetry and median grain size (D50) of the substrate, as well as the coordinates of known seagrass patches, on the performance of a Random Forest (RF) classifier used to determine seagrass cover. Using 148 Landsat images of the Venice Lagoon (Italy) between 1999 and 2020, we trained a RF classifier with only spectral features from Landsat images and seagrass surveys, respectively from 2002 and 2017. Then, by adding the features above and applying a time-based correction on predictions, we created multiple RF models with different feature combinations. We tested the quality of the resulting seagrass cover predictions from each model against field surveys, showing that bathymetry, D50 and coordinates of known patches exert an influence that is dependant on the training Landsat image and seagrass survey chosen. In models trained on a survey from 2017, where using only spectral features causes predictions to overestimate seagrass surface area, no significant change in model performance was observed. Conversely, in models trained on a survey from 2002, the addition of the out-of-image features and particularly coordinates of known vegetated patches greatly improves the predictive capacity of the model, while still allowing the detection of seagrass beds absent in the reference field survey. Applying a time-based correction eliminates small temporal variations in predictions, improving predictions that performed well before correction. We conclude that accounting for the coordinates of known seagrass patches, together with applying a time-based correction, has the most potential to produce reliable frequent predictions of seagrass cover. While this case study alone is insufficient to explain how geographic location information influences the classification process, we suggest that it is linked to the inherent spatial auto-correlation of seagrass meadow distribution. In the interest of improving remote sensing classification and particularly to develop our capacity to map vegetation across time, we identify this phenomenon as warranting further research.
- 24 January 2023
- Published by Copernicus GmbH in EGUsphere
- Vol. 2023, 1-32
Open AccessToward coherent space-time mapping of seagrass cover from satellite data: example of a Mediterranean lagoon
Seagrass meadows are a highly productive and economically important shallow coastal habitat. Their sensitivity to natural and anthropogenic disturbances, combined with their importance for local biodiversity, carbon stocks and sediment dynamics, motivate a frequent monitoring of their distribution. However, generating time-series of seagrass cover from field observations is costly, and mapping methods based on remote sensing require restrictive conditions on seabed visibility, limiting the frequency of observations. In this contribution, we examine the effect of accounting for environmental factors such as the bathymetry and median grain size (D50) of the substrate, as well as the coordinates of known seagrass patches, on the performance of a Random Forest (RF) classifier used to determine seagrass cover. Using 148 Landsat images of the Venice Lagoon (Italy) between 1999 and 2020, we trained a RF classifier with only spectral features from Landsat images and seagrass surveys, respectively from 2002 and 2017. Then, by adding the features above and applying a time-based correction on predictions, we created multiple RF models with different feature combinations. We tested the quality of the resulting seagrass cover predictions from each model against field surveys, showing that bathymetry, D50 and coordinates of known patches exert an influence that is dependant on the training Landsat image and seagrass survey chosen. In models trained on a survey from 2017, where using only spectral features causes predictions to overestimate seagrass surface area, no significant change in model performance was observed. Conversely, in models trained on a survey from 2002, the addition of the out-of-image features and particularly coordinates of known vegetated patches greatly improves the predictive capacity of the model, while still allowing the detection of seagrass beds absent in the reference field survey. Applying a time-based correction eliminates small temporal variations in predictions, improving predictions that performed well before correction. We conclude that accounting for the coordinates of known seagrass patches, together with applying a time-based correction, has the most potential to produce reliable frequent predictions of seagrass cover. While this case study alone is insufficient to explain how geographic location information influences the classification process, we suggest that it is linked to the inherent spatial auto-correlation of seagrass meadow distribution. In the interest of improving remote sensing classification and particularly to develop our capacity to map vegetation across time, we identify this phenomenon as warranting further research.
- Open Access
- 28 March 2022
- Published by Copernicus GmbH
High frequency spatio-temporal dynamics of seagrass meadows in a mediterranean lagoon
Seagrass meadows are a famously productive habitat found globally in intertidal and shallow subtidal shelves, hosting a rich biodiversity and efficiently sequestering carbon. They also exert an influence on coastal morphodynamics by modifying tidal flat roughness and erodibility, affecting sediment dynamics in coastal environments in a way that is not yet fully documented.Due to their sensitivity to environmental change and anthropogenic pressure, seagrass is prone to widespread die-off, which can be worsened by punctual degradation events such as dredging. Conversely, high primary productivity allows degraded meadows to recover rapidly, as well as expand through clonal and sexual reproduction. This potential for rapid change, however, is not currently matched by the frequency of observations of seagrass meadows, making it difficult to assess the spatial and temporal dynamics of seagrass meadows and their impact on intertidal zones.Using a novel method combining machine learning and time-series analysis, we extract seasonal maps of seagrass cover in the Southern Venice Lagoon from over 150 Landsat images over the 1999-2021 period and over 100 Sentinel images over the 2017-2021 period. By analysing changes in seagrass distribution over time, we identify seasonal extrema in seagrass surface area and observe their decadal evolution. Furthermore, we record the frequency and magnitude of sudden seagrass die-off events as well as recovery times. From these data we identify regions of preserved root mat, and by coupling seagrass dynamics with additional environmental data such as water temperature, sediment concentration and wave height, propose a set of possible degradation drivers and conditions for seagrass meadow recovery.With this contribution, we show how high-frequency mapping of seagrass distribution can reveal spatio-temporal dynamics of a highly productive coastal ecosystem, as well as offer keys to their response to environmental change.
- 28 March 2022
- Published by Copernicus GmbH
Open AccessHigh frequency spatio-temporal dynamics of seagrass meadows in a mediterranean lagoon
Seagrass meadows are a famously productive habitat found globally in intertidal and shallow subtidal shelves, hosting a rich biodiversity and efficiently sequestering carbon. They also exert an influence on coastal morphodynamics by modifying tidal flat roughness and erodibility, affecting sediment dynamics in coastal environments in a way that is not yet fully documented.Due to their sensitivity to environmental change and anthropogenic pressure, seagrass is prone to widespread die-off, which can be worsened by punctual degradation events such as dredging. Conversely, high primary productivity allows degraded meadows to recover rapidly, as well as expand through clonal and sexual reproduction. This potential for rapid change, however, is not currently matched by the frequency of observations of seagrass meadows, making it difficult to assess the spatial and temporal dynamics of seagrass meadows and their impact on intertidal zones.Using a novel method combining machine learning and time-series analysis, we extract seasonal maps of seagrass cover in the Southern Venice Lagoon from over 150 Landsat images over the 1999-2021 period and over 100 Sentinel images over the 2017-2021 period. By analysing changes in seagrass distribution over time, we identify seasonal extrema in seagrass surface area and observe their decadal evolution. Furthermore, we record the frequency and magnitude of sudden seagrass die-off events as well as recovery times. From these data we identify regions of preserved root mat, and by coupling seagrass dynamics with additional environmental data such as water temperature, sediment concentration and wave height, propose a set of possible degradation drivers and conditions for seagrass meadow recovery.With this contribution, we show how high-frequency mapping of seagrass distribution can reveal spatio-temporal dynamics of a highly productive coastal ecosystem, as well as offer keys to their response to environmental change.
- Open Access
- 4 March 2021
- Published by Copernicus GmbH
Long-term seasonal dynamics of seagrass extent in a Mediterranean Lagoon (Venice, Italy) from public satellite data
Seagrass meadows are globally distributed ecosystems found on coastal shelves, where they typically occupy the intertidal and shallow subtidal zones. Their sensitivity to environmental changes and proximity to human activities puts them at risk of degradation; as highly valuable marine habitats and potent carbon sequestration agents, they are therefore the target of numerous conservation programmes.The presence of seagrass strongly influences both wave and current propagation as well as sediment mobility, affecting the morphodynamics of entire estuarine systems. However, quantifying this influence is difficult, because our knowledge of seagrass cover is limited by its dynamic nature. Sensitivity to environmental factors such as nutrient load, available sunlight (mediated by turbidity) and temperature makes seagrass meadows prone to widespread changes in extent and density. Further degradation may occur stochastically through fishing and aquaculture. Conversely, seasonal cycles of high productivity allow meadows to recover and colonize new grounds through clonal and sexual reproduction. We propose a novel and cost-effective method to monitor seagrass cover in shallow waters across its seasonal and interannual variations. Combining machine learning and simple numerical modelling, we create a dense time-series of seagrass extent using over 100 LandSat scenes covering a 20-year-long period in the Venice Lagoon, Italy. Based on an expert-lead ecological survey (2004), we train one binary Random Forest Classifier in each of 5 environmentally-homogeneous geographical subzones, using spectral reflectance in the blue, green, red and near-infrared bands of the corresponding LandSat scene as recognition features. We then predict seagrass presence probability for LandSat scenes spanning the 1999-2019 period. Such predictions are made unstable by their sensitivity to sediment plumes as well as algal or gelbstoff blooms. Classification is therefore constrained by a simple numerical model that simulates clonal and sexual reproduction, regional die-off and punctual degradation. The model examines the potential areas colonised or degraded with respect to previous scenes, iteratively building stabilised maps of seagrass cover over time.Results are verified using further expert-lead estival surveys of the lagoon (2009, 2010, 2017) and of the inlets of Lido, Malamocco and Chioggia (2006 to 2015), as well as 10 digitised seagrass patches for a subsample of 20 invernal scenes. Accuracy metrics improve on the raw predictions (>80% to >85%), and scene-to-scene variability is reduced(>50% to <5%). These results show that public satellite data can be used to map seagrass cover and monitor its seasonal variations. In the future, cover maps may be used to estimate carbon storage, improve sediment transport models in shallow coastal areas, or identify drivers of change in seagrass meadows.
- 4 March 2021
- Published by Copernicus GmbH
Open AccessLong-term seasonal dynamics of seagrass extent in a Mediterranean Lagoon (Venice, Italy) from public satellite data
Seagrass meadows are globally distributed ecosystems found on coastal shelves, where they typically occupy the intertidal and shallow subtidal zones. Their sensitivity to environmental changes and proximity to human activities puts them at risk of degradation; as highly valuable marine habitats and potent carbon sequestration agents, they are therefore the target of numerous conservation programmes.The presence of seagrass strongly influences both wave and current propagation as well as sediment mobility, affecting the morphodynamics of entire estuarine systems. However, quantifying this influence is difficult, because our knowledge of seagrass cover is limited by its dynamic nature. Sensitivity to environmental factors such as nutrient load, available sunlight (mediated by turbidity) and temperature makes seagrass meadows prone to widespread changes in extent and density. Further degradation may occur stochastically through fishing and aquaculture. Conversely, seasonal cycles of high productivity allow meadows to recover and colonize new grounds through clonal and sexual reproduction. We propose a novel and cost-effective method to monitor seagrass cover in shallow waters across its seasonal and interannual variations. Combining machine learning and simple numerical modelling, we create a dense time-series of seagrass extent using over 100 LandSat scenes covering a 20-year-long period in the Venice Lagoon, Italy. Based on an expert-lead ecological survey (2004), we train one binary Random Forest Classifier in each of 5 environmentally-homogeneous geographical subzones, using spectral reflectance in the blue, green, red and near-infrared bands of the corresponding LandSat scene as recognition features. We then predict seagrass presence probability for LandSat scenes spanning the 1999-2019 period. Such predictions are made unstable by their sensitivity to sediment plumes as well as algal or gelbstoff blooms. Classification is therefore constrained by a simple numerical model that simulates clonal and sexual reproduction, regional die-off and punctual degradation. The model examines the potential areas colonised or degraded with respect to previous scenes, iteratively building stabilised maps of seagrass cover over time.Results are verified using further expert-lead estival surveys of the lagoon (2009, 2010, 2017) and of the inlets of Lido, Malamocco and Chioggia (2006 to 2015), as well as 10 digitised seagrass patches for a subsample of 20 invernal scenes. Accuracy metrics improve on the raw predictions (>80% to >85%), and scene-to-scene variability is reduced(>50% to <5%). These results show that public satellite data can be used to map seagrass cover and monitor its seasonal variations. In the future, cover maps may be used to estimate carbon storage, improve sediment transport models in shallow coastal areas, or identify drivers of change in seagrass meadows.
- Open Access
- 18 December 2019
- Published by MDPI AG in Remote Sensing
- Vol. 12 (1) , 13
Detecting the Morphology of Prograding and Retreating Marsh Margins—Example of a Mega-Tidal Bay
Retreat and progradation make the edges of salt marsh platforms their most active features. If we have a single topographic snapshot of a marsh, is it possible to tell if some areas have retreated or prograded recently or if they are likely to do so in the future? We explore these questions by characterising marsh edge topography in mega-tidal Moricambe Bay (UK) in 2009, 2013 and 2017. We first map outlines of marsh platform edges based on lidar data and from these we generate transverse topographic profiles of the marsh edge 10 m long and 20 m apart. By associating profiles with individual retreat or progradation events, we find that they produce distinct profiles when grouped by change event, regardless of event magnitude. Progradation profiles have a shallow scarp and low relief that decreases with event magnitude, facilitating more progradation. Conversely, steep-scarped, high-relief retreat profiles dip landward as retreat reveals older platforms. Furthermore, vertical accretion of the marsh edge is controlled by elevation rather than its lateral motion, suggesting an even distribution of deposition that would allow bay infilling were it not limited by the migration of creeks. While we demonstrate that marsh edges can be quantified with currently available DTMs, oblique observations are crucial to fully describe scarps and better inform their sensitivity to wave and current erosion.
- 18 December 2019
- Published by MDPI AG in Remote Sensing
- Vol. 12 (1) , 13
Open AccessDetecting the Morphology of Prograding and Retreating Marsh Margins—Example of a Mega-Tidal Bay
Retreat and progradation make the edges of salt marsh platforms their most active features. If we have a single topographic snapshot of a marsh, is it possible to tell if some areas have retreated or prograded recently or if they are likely to do so in the future? We explore these questions by characterising marsh edge topography in mega-tidal Moricambe Bay (UK) in 2009, 2013 and 2017. We first map outlines of marsh platform edges based on lidar data and from these we generate transverse topographic profiles of the marsh edge 10 m long and 20 m apart. By associating profiles with individual retreat or progradation events, we find that they produce distinct profiles when grouped by change event, regardless of event magnitude. Progradation profiles have a shallow scarp and low relief that decreases with event magnitude, facilitating more progradation. Conversely, steep-scarped, high-relief retreat profiles dip landward as retreat reveals older platforms. Furthermore, vertical accretion of the marsh edge is controlled by elevation rather than its lateral motion, suggesting an even distribution of deposition that would allow bay infilling were it not limited by the migration of creeks. While we demonstrate that marsh edges can be quantified with currently available DTMs, oblique observations are crucial to fully describe scarps and better inform their sensitivity to wave and current erosion.