Data Management Plan

These data were generated to provide insight into traffic patterns on a macro scale so they could be analyzed across the coastal waters of the Continental United States. For this dataset, a transit is counted for every unique vessel intersecting a 1 kilometer square grid cell each day. This data represents the total number of vessel transits from October 2009 - October 2010. There were some grid cells which were unable to be processed, but it is not perceived that this interferes with the integrity of these data.

Please note multiple connection errors occurred during the time frame of this study. In most cases, data gaps were filled by making subsequent requests to the Coast Guard or other groups receiving the same data feed. However, due to resource constraints, uninterrupted coverage was not obtained. Overall data outages were minimal, on the order of less than a day per month. Because outages were random and affect all areas uniformly, they do not have a significant effect on the integrity of the data. As stated on the USCG NAIS website, AIS data are not representative of all vessel traffic and USCG NAIS receivers do not fully cover the entire extent of this study area. Please take time to understand both of these limitations.

Process Steps:

• 2012-07-01 00:00:00 - Source data derived from the raw AIS data processing is as follows: The USCG maintains a network of AIS receivers that collects AIS messages from passing ships. These data are transmitted to USCG data center that compiles the data and provides data feeds to other government agencies. In accordance with the USCG COMDTINST 5230.80, the USCG provided OCS with a "Level A" data feed. Level A is unfiltered real-time data that is less than 96 hours from initial time of transmission. OCS has subscribed to this data since 2008. To limit data storage requirements the data feed was filtered by the USCG to only send one position message per ship per minute and all duplicate messages (i.e. ship broadcasts received by more than one NAIS station) were removed. This real time feed was archived at OCS. A specialized software NOAADATA.py (K. Schwehr. The noaadata-py Software Tool-set, v0.42, 2009. http://vislab-ccom.-unh.edu/schwehr/software/noaadata) is used to create daily files and and load the data into an Oracle Spatial database. Despite receiving this filtered feed, a great deal of conditioning needs to take place to prepare this data to be analyzed. The AIS system was not intended or designed for subsequent analysis; however Calder and Schwehr ably proved, given the proper conditioning, AIS data can be used for traffic analysis. ( B. R. Calder., K. Schwehr. Traffic Analysis for the Calibration of Risk Assessment Methods. Proceedings: US Hydrographic Conference 2009, Norfolk, VA, 11-14 May 2009, http://www.thsoa.org/us09papers.htm) Based on their research, we filtered out AIS messages with non-unique user IDs and vessels with erroneous dimensions. We also separated messages by speed, separating those reporting a speed of less than 0.4 knots into a separate anchored table. The accuracy and abundance of the AIS data support high resolution analysis specifically within port areas. This study however was interested in traffic patterns on a macro scale so patterns could be analyzed across the coastal waters of the Continental United States. To limit the processing time and the overall file size for each region a grid cell size of 1 kilometer was chosen. For this dataset a transits is counted for every unique vessel intersecting a grid cell each day. Multiple trips into a grid cell on a given day by the same vessel are only counted as one transit. Instead of calculating transits based on the coordinates within the AIS message transit lines were created by connecting all the vessels reports each day. Traffic counts were then calculated by summing the number of lines within each cell. Although this dataset only contains the total traffic count "TRNSTS_TTL" the original dataset has transits by AIS vessel type and many other attributes as well. Processed from May - July 2012. (Citation: NATIONWIDE AUTOMATIC IDENTIFICATION SYSTEM)
• 2012-08-01 00:00:00 - Further data development processing followed to obtain this feature class: Acquire original source data and maintain a copy on the D-tier (data01_d), \\csc-s-san1b\marinsp\MarineCadastre\original\2012\OCS AIS GRIDS. Created a geodatabase in a person workspace and imported the original shapefiles into this file geodatabase format. Added a field in each of the feature classes to base classification ranges on, deleted all values of 0 in each of the data sets. Using the field calculator populated the new classification field values based on selecting the values of 'TRNSTS_TTL' field - performed this on each UTM zone feature class. Then performed a dissolve on each of the individual UTM zones datasets based on the classification field. Created a new feature classes with NAD 83 projection, added a field for the classification code, and loaded the data from the classified dissolved dataset the new NAD 83 dataset. Validate the feature classes' geometry (Data Management > Features > Repair Geometry). Added field in each of the three new data sets called 'vesselCountRange', populated this new field using field calculator to display the actual range of values within that classification code. If necessary, open ArcMap and in a new MXD click File > Map Document Properties and check the box "Store relative path names to data sources". Click OK. Added in the Oceans Basemap in the working MXD to work up the color scheme for the data. Select the fields that will be visible (chose to hide Classification Code and show Vessel Count Range). Symbolize the layer by Classification Code field to represent the Vessel Count Ranges. Develop a layer file containing symbology from each feature class. Include Layer Name, Description, Credits, and Scale Ranges (if applicable) within the layer file and its source. Arrange quality process review by peers. Review quality process feedback, determine applicability, and make adjustments to data to meet process specifications. Place a copy of the feature class/es within SDE. Project the symbolized layer into Web Mercator and place within the viewer's MXD. Work with the MXD steward to determine where and how the layer fits within the MXD. Acquire, create and update metadata from providers (as needed). Annually evaluate source data to determine requirement for replacement. (Citation: AIS Vessel Count - 2010)