We are analyzing our LIS Bird Count data, from late 2013, and planning how best to collect data in late 2014. We receive and store data of various types, including...
- precise locational data from eBird,
- precise locational data from ship transects through LIS Long Island Sound,
- approximate locational data from ferries which cross LIS,
- data from sea watches from points along the shore, and
- data from somewhere within the 15-mile diameter circles from almost a dozen CBC Christmas Bird Counts.
A data point may be as precise as three Common Loons seen at a exact latitude/longitude point, or as approximate as three Common Loons seen (somewhere) in the Napatree CBC (a 15-mile circle). this is a lot "variability" (heterogeneity) in our 2013 data. Next winter can we find a better way to standardize all/most our data on some locational/spatial "common-denominator?"
We are looking for good way(s) to capture, store, analyze, and display this data, both for our current 2013 data, and for future years.
Ideally we would capture every observation at an exact lat/long point, and store it in our database that way, for whatever future aggregation and display we would like (such as summarizing in a grid, with bins of whatever size was wanted, and eventually displaying this as colored grids or as a "heat map").
Some bird observations will come to us with approximate locations, and we need to decide how to represent them in our database, perhaps as points which we know to be (and we should indicate to be) approximations of the actual observation location.
The questions/issues we face include the following:
1. Online data entry (via eBird, BirdLog/eBird, or some customized tool) and/or paper-based map data entry?
2. Points, lines, and/or polygons?
3. And we have some data from some long (50-mile) transects, which raises the possible need to divide each long transect down into multiple short (perhaps one kilometer or quarter-of-an-hour?) transects, perhaps with each small sub-transect having all its observation data anchored at its starting point, as was done with seabird transects off the Alaskan coast and the MBO data.
4. Regarding the paper-based maps, how do you digitize them into an online tool (or otherwise get them into the database)? By eyeball approximations, or by geo-referencing/calibration of some kind?
5. And how do you convert sketched areas (on paper maps, or via on online interface) into one point (or several points), by their approximate centers?
Researching these items is slowing down our compilations of our 2013 data, but that will be available by May, with caveats explained.
=======================================
RELATED WORK FOLLOWS:
Other spatial research work may provide some guidance for us. We are looking at the following efforts:
NPPSD. North Pacific Pelagic Seabird Database [40+ years].
North Pacific Pelagic Seabird Database (NPPSD) - USGS Alaska Science Center
A very large number of seabird trips and transect data.
Atlantic Canada Seabird Data (much of it from Dick Brown plus MBO Manomet Bird Observatory trips, Kevin Powers, et al) [1965-1992+]
OBIS-SEAMAP Dataset - PIROP Northwest Atlantic 1965-1992
Welcome to OBIS! | OBIS
ACJV Atlantic Coast Joint Venture [current].
A cooperative of several seabird research groups and projects.
Marine Birds
Northeast Ocean Data organization [the last few years]
About | Northeast Ocean Data Portal
Unfortunately this resource contains almost no information about seabirds currently.
Northeast Ocean Data Portal...
NOAA/Veit/Martin seabird censusing [tbd].
Forthcoming Ph.D. thesis of M.Martin.
Little information currently available.
Stellwagen Bank seabird censusing.
Center for Coastal Monitoring and Assessment: An Ecological Characterization of the Stellwagen
Bank NMS Region .
This excellent study reflects the MBO pelagic trips mostly from the 1980s and covering all of the GOM Gulf of Maine.
"Seabird observer methods followed Powers (1982). Observations were recorded in 15 minute periods, where each period was considered an individual transect and the spatial coordinates were recorded using the ship’s instruments. Seabirds were counted within a fixed strip width of 300 m at one side of the ship, traveling on a straight course, at a constant speed (approx. 8 - 12 knots). The duration of each sampling period was fixed, but transect length varied slightly depending on vessel speed. "
"NOAA’s Northeast Fisheries Science Center, Woods Hole, provided the MBO database in ASCII format. In addition, a version of the original was also integrated within the PIROP (Intégré des Recherches sur les Oiseaux Pélagiques) database held by OBIS-SEAMAP (Read et al., 2006). Although a small spatial overlap exists, the PIROP data were collected primarily for Canadian seabird monitoring purposes and focused mainly on the Canadian Atlantic, while the Manomet data focused on the U.S. continental shelf extending from Nova Scotia, Canada, south to Cape Hatteras, North Carolina. The predictive modelling component of this study used the combined PIROP and MBO database."
Stellwagen Bank National Marine Sanctuary
The new S4 program (A.M.Runfola) is underway but few results published thus far.
Area of focus limited to Stellwagen Bank National Marine Sanctuary and nearby...
LISP: Long Island Sound Project [2013 thru current].
That is THIS project, and will be summarized here, at this website, shortly.
We are trying to be as consistent as possible with these other efforts.
 |
| Standard eastern route, without H6 station excluded. |
 |
| Ten-minute arc squares, and CBC Christmas Bird Count circles. |
LISS: Long Island Sound Study [the last fifteen years].
Long Island Sound Study
Long Island Sound - Wikipedia, the free encyclopedia
The LISS does not focus on bird locations in the offshore waters of the LIS (the "mid-Sound"), but rather along the coast, beaches and marshes of the LIS. Could not find a portal for data or maps, and here are the maps which are available...
Maps « Long Island Sound Study
Long Island Sound Resource & Use Inventory, and Blue Plan [2015 forward].
This is a proposed new piece of work, awaiting approval to move forward.
....
TNC The Nature Conservancy is doing a "LISEA" LIS Ecological Assessment project which should provide some data into this new inventory. LISEA information and map will be published at this site in the near future...
Long Island Sound LISEA
New York State Offshore Spatial Plan [recent decades].
Includes more than a decade of seabird data from MBO the Manomet Bird Observatory, with many thanks to Kevin Powers.
But very minimal seabird coverage of Long Island Sound.
docs.dos.ny.gov/communitieswaterfronts/ocean_docs/NYSDOS_Offshore_Atlantic_Ocean_Study.pdf
 |
| You can click on any image to enlarge, and then hit ESC key (or browser back arrow) to return |
Mid-Atlantic waterbird survey, shown in the MARCO Mid-Atlantic Ocean Data Portal. [yrs tbd]
"All waterbird species in the USFWS database from winter surveys only were exported to density blocks". There is little or no information on seabird species thus far.
MARCO Marine Life
MARCO Marine Planner
NOAA work in Texas [current].
More information coming soon.
RNBWS [the last 50+ years]:
Royal Naval Birdwatching Society.
More information on the way.
eBird [the last ten years]
Pelagic birding in eBird | eBird
eBird Pelagic Protocol | eBird
Note that on April 15, 2014 eBird announced its "pelagic birding protocol" for birding two or more miles offshore, adopting the method used by most pelagic birding surveys for a long time (e.g., those in the Bering Sea and throughout the north Pacific, plus the MBO work on the US east coast).... divide long transects into short segments, and report the birds seen in each segment, and use the location at the beginning of the segment. eBird recommends an hour or less for each segment, although some very scientific surveys use very short segments, such as one minute or less (as used in the Stellwagen Bank surveys currently). eBird says,
"The protocol is a traveling count of 1 hour or less. Do your best to record (using a GPS or BirdLog) or estimate (using average ship speed) the distance traveled, and try to account for potential backtracking or changes in speed.... At the end of each count period, start a new checklist and plot your point at the beginning of each transect. It is strongly recommended to put water temperature and depth in the checklist comments, along with visibility and wave height, as these are important variables. Repeat this process throughout the trip until you return to within two miles of shore."
We totally agree in principle with this eBird protocol, although with much shorter time segments, especially since Long Island Sound is relatively small, and needs more precise locations than would a bird seen twenty-two miles off Cape Hatteras.
Of course the most precise data would be best: recording (and storing in the database) the exact time/lat/long of every bird seen, especially because it could be summarized in any possible way for future use. But it is much easier in the field to aggregate observed data in a way which reduces the burden on the observer (and their recording person). For our LIS Project we need to agree on how much precision is enough. Of course it would be great to capture the exact latitude/longitude of every bird we see in LISound, but that is impractical I think, because of the labor burden required to do that. So what is the best balance, and what are the mechanics we should use?
