Using Unmanned Aerial Systems (UAS) Capabilities to Help Identify Giant Salvinia (Salvinia molesta) and the effects of Salvinia weevil (Cyrtogagous salviniae) Active
USGS is collecting remotely sensed data to classify Salvinia molesta, a non-native aquatic species that reduces light and oxygen levels in water, making it unsuitable for fish and other aquatic plant life.
The Science Issue and Relevance: States along coastal waters rely on waterways, canals, and bayous as a main source of travel for recreation and marine industries. More and more often these waterways are filled with Salvinia, which is a floating aquatic weed that chokes channels with a thick mat of vegetation. This floating aquatic grows rapidly over the entire surface and can be a nuisance for fishing, swimming, and boating. This species of Salvinia has one of the highest documented growth rates among plants, and under perfect conditions can double its biomass in just 3 days. Salvinia molesta is a non-native plant that can change oxygen levels of the water, making it unsuitable for fish and other aquatic plant life. Once giant Salvinia becomes established, it is almost impossible to completely remove. One biological control program that is being used to have an impact on the Salvinia vegetation is by the introduction of the Salvinia weevil (Cyrtobagous salviniae). At high densities, weevil feeding on these plants cause the mats to turn brown and eventually die.
Methodology for Addressing Issue: The 3DR Solo Unmanned Aerial Systems (UAS), a quadrocopter designed to carry a single sensor and/or camera to capture remotely sensed data, was used to fly this mission. Automated flight plans were developed for each of the six study areas. For continuous and complete coverage, a flight plan with a 60% end lap and 60% side lap was created. Side and end laps, which refer to the amount of image overlap between photos, reduce the chance for missing imagery coverage. The altitude for this mission was 150 feet above ground level. At this elevation a pixel resolution or Ground Sample Distance (GSD) of 0.5 inches is achieved. To acquire the imagery, the Ricoh GR II digital single lens camera captured individual frames at 2-second intervals for each of the study areas. These frames were used to create a seamless orthomosaic using the red, green, and blue (RGB) visible light spectrum (natural color). The MicaSense RedEdge multispectral sensor is also used to assist with the identification of vegetation species. The RedEdge has 5 bands which allows for plant health evaluation.
Future Steps: The imagery collected from the UAS will be analyzed to classify vegetation into 7 classes: (1) Land, (2) Water, (3) Healthy Salvinia, (4) Damaged Salvinia, (5) Dying Salvinia,(6) Dead Salvinia, and (7) Other Floating Vegetation.
USGS is collecting remotely sensed data to classify Salvinia molesta, a non-native aquatic species that reduces light and oxygen levels in water, making it unsuitable for fish and other aquatic plant life.
The Science Issue and Relevance: States along coastal waters rely on waterways, canals, and bayous as a main source of travel for recreation and marine industries. More and more often these waterways are filled with Salvinia, which is a floating aquatic weed that chokes channels with a thick mat of vegetation. This floating aquatic grows rapidly over the entire surface and can be a nuisance for fishing, swimming, and boating. This species of Salvinia has one of the highest documented growth rates among plants, and under perfect conditions can double its biomass in just 3 days. Salvinia molesta is a non-native plant that can change oxygen levels of the water, making it unsuitable for fish and other aquatic plant life. Once giant Salvinia becomes established, it is almost impossible to completely remove. One biological control program that is being used to have an impact on the Salvinia vegetation is by the introduction of the Salvinia weevil (Cyrtobagous salviniae). At high densities, weevil feeding on these plants cause the mats to turn brown and eventually die.
Methodology for Addressing Issue: The 3DR Solo Unmanned Aerial Systems (UAS), a quadrocopter designed to carry a single sensor and/or camera to capture remotely sensed data, was used to fly this mission. Automated flight plans were developed for each of the six study areas. For continuous and complete coverage, a flight plan with a 60% end lap and 60% side lap was created. Side and end laps, which refer to the amount of image overlap between photos, reduce the chance for missing imagery coverage. The altitude for this mission was 150 feet above ground level. At this elevation a pixel resolution or Ground Sample Distance (GSD) of 0.5 inches is achieved. To acquire the imagery, the Ricoh GR II digital single lens camera captured individual frames at 2-second intervals for each of the study areas. These frames were used to create a seamless orthomosaic using the red, green, and blue (RGB) visible light spectrum (natural color). The MicaSense RedEdge multispectral sensor is also used to assist with the identification of vegetation species. The RedEdge has 5 bands which allows for plant health evaluation.
Future Steps: The imagery collected from the UAS will be analyzed to classify vegetation into 7 classes: (1) Land, (2) Water, (3) Healthy Salvinia, (4) Damaged Salvinia, (5) Dying Salvinia,(6) Dead Salvinia, and (7) Other Floating Vegetation.