I’m planning a project that will assess the performance of side scan sonar, drone, fixed-wing manned aerial, and satellite imagery surveys in accurately detecting eelgrass meadow edge and low-density areas. All 4 remote survey methods, groundtruthed by diver surveys that will map the meadow to the last shoot, will be done at several edge transects across several different meadows. To reduce the influence of time on precise edge location and shoot density across surveys, the surveys must all take place reasonably close to one another in time. My thought is they should be within the same 2-4 week span. Is that reasonable, and short enough to be considered a negligible amount of time between survey methods?
I don’t have any experience with eelgrass so hopefully @dwilcox@brookelandry or @Madison_Lytle can chime in, but I would think if you are able to conduct the surveys around peak growth/biomass that time span should be reasonable. (but again I say that having never worked with eelgrass so I could be completely full of it).
Good luck and keep us posted on which methods works out best!
-Dottie
So long as there are no dramatic losses in the bed, such as a heat, cold, or wind event, I would not expect very much change in plant characteristics during a two-week time period. We found that the conditions when imagery was acquired have a major influence on the ability to detect the grass from aerial or satellite imagery. If the different types of imagery are aquired under significantly different conditions, it may not be an accurate comparison. We find that the number days where the water and skies are clear is pretty limited in Chesapeake Bay and we are lucky to be able to carefully target aerial imagery on those days. If your conditions are more stable, mapping will be easier as will the comparison.
Hi, Jill. Maine DEP and partners completed a similar methods assessment (more sonar methods, no satellite component) a few years ago in Casco Bay, Maine, and used a longer index period (~mid-June -mid-Sept.) over which to gather the data, but our work wasn’t as specifically focused on edge detection and low density beds. If the main metrics you are comparing are edge location and density, I would think completing all methods within 2-4 weeks would be ideal. If aboveground biomass is an important metric, I would expect closer to 2 weeks would be best for comparability. Really awesome that you are doing this work and I can’t wait to hear more about it!
Jill, the project coordination was overseen by Claire Enterline (Maine Coastal Program). She intends to produce a report on the outcomes, over the winter, I believe. If you don’t already know Claire, she can be reached at claire.enterline@maine.gov.
).