Using RADAR AIS to investigate ship behavior in the Chesapeake Bay ocean approach off of Virginia, USA

摘要

We used a combination of AIS and RADAR to characterize large (≥ 65' long) vessel traffic in the Chesapeake Bay ocean approach along the US eastern seaboard from May 2008 through April 2009. During the 60 days of monthly surveys, we recorded over 2.6×10⁶ records of data with ship position information. There were 1181 hours of AIS and 540 hours of RADAR data collected for a monthly mean of 98 and 45 hours for AIS and RADAR respectively. These data represented 1411 transits by vessels broadcasting AIS for a total of 69,606km of track line, and vessels acquired with RADAR represented a total of 506 transits for a total of 8,702km of track line. AIS and RADAR data resulted in 1.2 and 0.9 transits per hour respectively. We corrected the AIS data to match RADAR effort, and, when effort was equal, AIS data represented only 49.7% of the total large vessel transits observed, and transits per hour were equal for corrected AIS and RADAR data. We recorded more inbound vessels using AIS and more outbound vessels with RADAR. The density pattern of vessels recorded using AIS differed from the RADAR pattern. There was a very discrete pattern for vessels broadcasting an AIS signal that corresponded with the shipping lane buoys. Vessels recorded using RADAR, on the other hand, were more dispersed and displayed a less discrete pattern covering a large area outside of the ship buoys. Winter was the season with the fewest vessel transits for both data collection methods (23.8% AIS; 17.0% RADAR), though there was not a substantial decrease over other seasons in the AIS data. Spring, summer and fall were similar for AIS (24.2%, 25.7%, 26.3% respectively), but fall (31.4%) had considerably more vessels recorded by RADAR than spring and summer (26.5%, 25.1% respectively). When examining vessel type data, more than two thirds of the vessel transits recorded using AIS were cargo vessels (68%; n=337). Excluding unknown vessels, military vessels made up 47% (- =179) of the vessels we identified with RADAR, followed by fishing (32%; n=120) and federal law enforcement (USCG) vessels (10%, n=38). Spatial analysis was necessary to accurately evaluate ship speed over the entire study area. Spatial analyses of these data allowed us to examine speed behavior in a grid format, reducing the effect of directional bias from an overrepresentation of records from relatively slow moving vessels. Without utilizing spatial analysis, one could draw the conclusion that vessels were transiting the area more slowly than they actually were because AIS speed records are skewed by the greater number of speed points for slower vessels. There was an obvious reduction in speed of vessels transmitting AIS in the shipping lanes from before to after the Ship Strike Rule that was enacted in December 2008 to protect the western North Atlantic right whale (Eubalaena glacialis). The Ship Strike Rule identified a Seasonal Management Area (SMA) within which commercial ships ≥65 ft. in length were required to slow to 10kts or less for part of the year (50 CFR224.105). Although vessels slowed, many were not in compliance with the mandatory speed limit of the Rule. Speed restrictions had no effect on vessels observed using RADAR, most of which were not required to comply with the regulations. Cargo ships can be quite large and travel at speeds in excess of 20kt, and were required to broadcast AIS signals, so their behavior was relatively easy to assess. Few cargo vessels traveled 10kt or less for an entire transit through the SMA. Thus, we observed an apparent reaction to the speed restrictions in the SMA, but did not observe strict compliance. Recently the US Coast Guard has begun issuing speeding citations to vessels that have been recorded blatantly disregarding the speed restrictions in SMAs. We applaud these efforts and suspect that if the effort continues consistently it will encourage compliance. Our data suggest that A