| ArcGIS REST Services Directory |
| Home > services > SDM_Features (FeatureServer) > All Layers and Tables | | API Reference |
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presence-absence models had good predictive capacity with cross-validated Area Under the Receiver Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in the Newfoundland and Labrador region.
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presece-absence models had good predictive capacity with cross validated Area Under the Reciever Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in he Newfoundland and Labrador region.
This polygon shapefile represents the spatial extent of the training data or alternatively the area tha was not considered 'extrapolated' by the random forest model. This polygon was created by converting the data-present rasters provided in the source data to a polygon. Extrapolated area represents the regions where the modelling boundary extends far beyond the spatial extent of the training data. Extrapolation of model predictions to areas outside the range of data observations may produce unreliable predictions in those areas (Elith et al. 2010). The authors define areas of extrapolation as those areas where at least one environmental variable has values above or below its sampled range.
The associated document can be found:
https://waves-vagues.dfo-mpo.gc.ca/Library/40577806.pdf
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presece-absence models had good predictive capacity with cross validated Area Under the Reciever Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in he Newfoundland and Labrador region.
This polygon shapefile represents the spatial extent of the extrapolated area, without the data present area. Extrapolated area represents the regions where the modelling boundary extends far beyond the spatial extent of the training data. This polygon was created by erasing the data present polygon from the entire extrapolated area polygon. converting the extrapolated rasters provided in the source data to a polygon. Extrapolation of model predictions to areas outside the range of data observations may produce unreliable predictions in those areas (Elith et al. 2010). The authors define areas of extrapolation as those areas where at least one environmental variable has values above or below its sampled range.
The associated document can be found:
https://waves-vagues.dfo-mpo.gc.ca/Library/40577806.pdf
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presece-absence models had good predictive capacity with cross validated Area Under the Reciever Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in he Newfoundland and Labrador region.
This polygon shapefile represents entire area 'extrapolated' by the random forest model, area containing presence-absence data and area of extrapolated predictions. This polygon was created by converting the extrapolated rasters provided in the source data to a polygon. Extrapolated area represents the regions where the modelling boundary extends far beyond the spatial extent of the training data. Extrapolation of model predictions to areas outside the range of data observations may produce unreliable predictions in those areas (Elith et al. 2010). The authors define areas of extrapolation as those areas where at least one environmental variable has values above or below its sampled range.
The associated document can be found:
https://waves-vagues.dfo-mpo.gc.ca/Library/40577806.pdf
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presence-absence models had good predictive capacity with cross-validated Area Under the Receiver Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in the Newfoundland and Labrador region.
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presece-absence models had good predictive capacity with cross validated Area Under the Reciever Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in he Newfoundland and Labrador region.
This polygon shapefile represents the spatial extent of the training data or alternatively the area that was not considered 'extrapolated' by the random forest model. This polygon was created by converting the data-present rasters provided in the source data to a polygon. Extrapolated area represents the regions where the modelling boundary extends far beyond the spatial extent of the training data. Extrapolation of model predictions to areas outside the range of data observations may produce unreliable predictions in those areas (Elith et al. 2010). The authors define areas of extrapolation as those areas where at least one environmental variable has values above or below its sampled range.
The associated document can be found:
https://waves-vagues.dfo-mpo.gc.ca/Library/40577806.pdf
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presece-absence models had good predictive capacity with cross validated Area Under the Reciever Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in he Newfoundland and Labrador region.
This polygon shapefile represents the spatial extent of the extrapolated area, without the data present area. Extrapolated area represents the regions where the modelling boundary extends far beyond the spatial extent of the training data. This polygon was created by erasing the data present polygon from the entire extrapolated area polygon. converting the extrapolated rasters provided in the source data to a polygon. Extrapolation of model predictions to areas outside the range of data observations may produce unreliable predictions in those areas (Elith et al. 2010). The authors define areas of extrapolation as those areas where at least one environmental variable has values above or below its sampled range.
The associated document can be found:
https://waves-vagues.dfo-mpo.gc.ca/Library/40577806.pdf
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presece-absence models had good predictive capacity with cross validated Area Under the Reciever Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in he Newfoundland and Labrador region.
This polygon shapefile represents entire area 'extrapolated' by the random forest model, area containing presence-absence data and area of extrapolated predictions. This polygon was created by converting the extrapolated rasters provided in the source data to a polygon. Extrapolated area represents the regions where the modelling boundary extends far beyond the spatial extent of the training data. Extrapolation of model predictions to areas outside the range of data observations may produce unreliable predictions in those areas (Elith et al. 2010). The authors define areas of extrapolation as those areas where at least one environmental variable has values above or below its sampled range.
The associated document can be found:
https://waves-vagues.dfo-mpo.gc.ca/Library/40577806.pdf
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presence-absence models had good predictive capacity with cross-validated Area Under the Receiver Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in the Newfoundland and Labrador region.
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presece-absence models had good predictive capacity with cross validated Area Under the Reciever Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in he Newfoundland and Labrador region.
This polygon shapefile represents the spatial extent of the training data or alternatively the area that was not considered 'extrapolated' by the random forest model. This polygon was created by converting the data-present rasters provided in the source data to a polygon. Extrapolated area represents the regions where the modelling boundary extends far beyond the spatial extent of the training data. Extrapolation of model predictions to areas outside the range of data observations may produce unreliable predictions in those areas (Elith et al. 2010). The authors define areas of extrapolation as those areas where at least one environmental variable has values above or below its sampled range.
The associated document can be found:
https://waves-vagues.dfo-mpo.gc.ca/Library/40577806.pdf
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presece-absence models had good predictive capacity with cross validated Area Under the Reciever Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in he Newfoundland and Labrador region.
This polygon shapefile represents the spatial extent of the extrapolated area, without the data present area. Extrapolated area represents the regions where the modelling boundary extends far beyond the spatial extent of the training data. This polygon was created by erasing the data present polygon from the entire extrapolated area polygon. converting the extrapolated rasters provided in the source data to a polygon. Extrapolation of model predictions to areas outside the range of data observations may produce unreliable predictions in those areas (Elith et al. 2010). The authors define areas of extrapolation as those areas where at least one environmental variable has values above or below its sampled range.
The associated document can be found:
https://waves-vagues.dfo-mpo.gc.ca/Library/40577806.pdf
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presece-absence models had good predictive capacity with cross validated Area Under the Reciever Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in he Newfoundland and Labrador region.
This polygon shapefile represents entire area 'extrapolated' by the random forest model, area containing presence-absence data and area of extrapolated predictions. This polygon was created by converting the extrapolated rasters provided in the source data to a polygon. Extrapolated area represents the regions where the modelling boundary extends far beyond the spatial extent of the training data. Extrapolation of model predictions to areas outside the range of data observations may produce unreliable predictions in those areas (Elith et al. 2010). The authors define areas of extrapolation as those areas where at least one environmental variable has values above or below its sampled range.
The associated document can be found:
https://waves-vagues.dfo-mpo.gc.ca/Library/40577806.pdf
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presece-absence models had good predictive capacity with cross validated Area Under the Reciever Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in he Newfoundland and Labrador region.
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presece-absence models had good predictive capacity with cross validated Area Under the Reciever Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in he Newfoundland and Labrador region.
This polygon shapefile represents the spatial extent of the training data or alternatively the area that was not considered 'extrapolated' by the random forest model. This polygon was created by converting the data-present rasters provided in the source data to a polygon. Extrapolated area represents the regions where the modelling boundary extends far beyond the spatial extent of the training data. Extrapolation of model predictions to areas outside the range of data observations may produce unreliable predictions in those areas (Elith et al. 2010). The authors define areas of extrapolation as those areas where at least one environmental variable has values above or below its sampled range.
The associated document can be found:
https://waves-vagues.dfo-mpo.gc.ca/Library/40577806.pdf
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presece-absence models had good predictive capacity with cross validated Area Under the Reciever Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in he Newfoundland and Labrador region.
This polygon shapefile represents the spatial extent of the extrapolated area, without the data present area. Extrapolated area represents the regions where the modelling boundary extends far beyond the spatial extent of the training data. This polygon was created by erasing the data present polygon from the entire extrapolated area polygon. converting the extrapolated rasters provided in the source data to a polygon. Extrapolation of model predictions to areas outside the range of data observations may produce unreliable predictions in those areas (Elith et al. 2010). The authors define areas of extrapolation as those areas where at least one environmental variable has values above or below its sampled range.
The associated document can be found:
https://waves-vagues.dfo-mpo.gc.ca/Library/40577806.pdf
Species distribution modelling using a random forest (RF) machine learning approach was used to predict the probability of occurrence of sponges, sea pens, and large and small gorgonian corals in the Newfoundland and Labrador Region. A suite of 66 environmental predictor variables from different data sources were used. Models utilized catch records from the DFO multispecies trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys, DFO/industry northern shrimp surveys and Spanish groundfish trawl surveys. Most presece-absence models had good predictive capacity with cross validated Area Under the Reciever Operating Characteristic Curve (AUC) values ranging from 0.786 to 0.926. These models were used in a Canadian Science Advisory Secretariat (CSAS) process to delineate significant areas of cold-water corals and sponges in he Newfoundland and Labrador region.
This polygon shapefile represents entire area 'extrapolated' by the random forest model, area containing presence-absence data and area of extrapolated predictions. This polygon was created by converting the extrapolated rasters provided in the source data to a polygon. Extrapolated area represents the regions where the modelling boundary extends far beyond the spatial extent of the training data. Extrapolation of model predictions to areas outside the range of data observations may produce unreliable predictions in those areas (Elith et al. 2010). The authors define areas of extrapolation as those areas where at least one environmental variable has values above or below its sampled range.
The associated document can be found:
https://waves-vagues.dfo-mpo.gc.ca/Library/40577806.pdf