NOAA Award Number NA11NOS0120038 (June 2011 – May 2016)
The Mid-Atlantic Regional Association Coastal Ocean Observing System (MARACOOS) continues to implement its partnership-based strategy to support stakeholder needs through sustained regional ocean observation and prediction in the Mid-Atlantic Bight (MAB) from Cape Hatteras to Cape Cod. Over the 5-year project duration, MARACOOS has (1) united and integrated the organizational activities of MACOORA (established in 2004) and the operational activities of MARCOOS (established in 2007) into the single entity known as MARACOOS; (2) maintained and expanded the existing observing, data management and forecasting subsystems focused on the transition from data-generated to model-generated ensemble ocean forecast products that target multiple users; and (3) expanded end-to-end operations across all five regional themes through (a) enhanced education and engagement activities, (b) the leveraging of resources beyond IOOS, and (c) the application of IOOS Association-endorsed metrics to measure and demonstrate success. During this period, MARACOOS has continued compiling relevant procedures and documenting other requirements toward certification as per 15 CFR Part 997, with a goal of submitting the documentation during summer 2016. MARACOOS requested and was granted a 1 year no-cost extension for this project through May 31, 2017.
PROGRESS AND ACCOMPLISHMENTS
MARACOOS has exploited both successes and lessons learned in moving forward toward its 5-year goals. Chief among these goals are the aim to expand from observation to predictions and the aim to expand the suite of valued information products across the 5 MARACOOS themes. In order to achieve these goals and thereby realize the potential envisioned in the 5-year plan, we see increasing user engagement and expanding leveraging of new and existing partnerships as crucial. Simultaneously, traditional performance metrics are being refined to gauge the spatial coverage and temporal reliability of the MARACOOS observations, and new metrics are being developed to assess the accuracy of these observations and our prediction products. Our milestone schedule is found on page 13, Section E of the MARACOOS proposal. In this report, progress towards milestones and metrics for each of the 5 subsystems are outlined first. In sections A-K, the work accomplished toward those milestones is discussed in greater detail.
Milestones and Metrics: Management Subsystem. During this reporting period, bi-weekly board meetings and operations meetings were held via conference call. The bi-weekly operations meetings included discussions on the Weather Prediction Ensemble Validation, Satellites, HF-Radar, Gliders, DMAC, Education & Outreach, Ocean Predictions, Fisheries, CBIBS, and QARTOD. Observing Subsystems: a) Weather: The MARACOOS weather ensemble operations have been run four times per day at 00, 06, 12 and 18 UTC out to 36 hours. WeatherFlow has been utilizing up to 11 prediction models in the ensemble. Ensemble and individual member performance has been quantified for 2015, and some results presented in the next section. b) Satellites: There were 0 days of data acquisition gaps for the University of Delaware L-band dish and Geo-Stationary Dish. The Geo-Stationary satellite receiving dish at the University of Delaware (UDel) receives 40-80 measurements of sea surface temperature (4km resolution) in the MARACOOS region daily. The Rutgers University L-band system failed after 23 years of operations. Both the Rutgers and University of Delaware X-Band dishes are non-operational and expensive repairs are needed. The X-band data stream is updated through our processing of raw data from NASA. c) HF-Radar: The MARACOOS metric for long range data coverage and availability has surpassed the 80/80 metric goal (80% coverage 80% of the time) again during this period. d) Gliders: MARACOOS glider operators deployed 11 gliders in the MARACOOS region during June 2015-May 2016, totaling 232 days and 5143 kilometers sampled. e) DMAC Subsystem: The uptime for observations and model data in the MARACOOS Data Center continues to be >99% and the auto-monitoring system provides feedback on most of the data feeds. f) Modeling Subsystem: ESPreSSO delivered on-time forecasts on all but 2 days, on both occasions due to corruption of the river inflow data file. Steps have been taken to harden processing against similar failures. Several data streams have become less reliable over the last 6 months, notably AVHRR SST via MARACOOS THREDDS. Since 1 January, 29 cycles were unable to download AVHRR on schedule. MARACOOS operations were notified on each occasion and data restored within 48 hours; so they eventually entered the final 3-day analysis cycle. On 2 occasions up to 4 days of corrupted 6 km CODAR files were identified by the ESPreSSO QC systems and MARACOOS operations were notified. Data were not restored in time to enter the real-time analysis. Throughout the reporting period recurring delays, or total absence of SST from WindSat and AMSR2 occurred because of on-board instrument problems. While data providers were notified, commonly there was no possible resolution and those cycles proceeded in the absence of microwave SST data. ESPreSSO access glider data from the internal Rutgers data center and as more of these data enter GTS we are adapting our QC procedure to try to avoid assimilating data duplicates. The Stevens NYHOPS models had the following metrics: a) deterministic 78hr Forecast Horizon (FH): 100% operational, b) 125-member hydrodynamic ensemble; 78hrs FH: 100% operational, c) 125-member Stevens SNAP-Ex extended ensemble; 105hrs FH: 100% operational, and d) 125-member Stevens Hydrologic ensemble for NYH contributing watersheds; 78hrs FH: 100%. The SMAST-HOPS model continued to make weekly nowcasts/predictions and was 100% available.
Meetings/Engagement Activity Highlights: Please see the Supplemental Report for this information.
A) Atmospheric Data Integration: For the MARACOOS domain, WeatherFlow has been generating a gridded ensemble of surface meteorological fields from a variety of available high-resolution mesoscale numerical weather predictions (NWP) (ensemble members). The ensemble products, which include the gridded data, production imagery, proprietary WeatherFlow observational data used in production and the statistics, have been available to the MARACOOS community.
The MARACOOS weather ensemble operations have been running four times per day at 00, 06, 12 and 18UTC out to 36 hours. WeatherFlow has been utilizing up to 11 prediction models in the ensemble. WeatherFlow has been taking advantage of available high-resolution (4km or better) NWP predictions generated by government, academic, and private sectors that cover all or significant portions of the MARACOOS domain. The list of models recently used in producing the ensemble is itemized in Table 1. As mentioned in previous reports, GRIB2 files are generated and available for anyone who may need atmospheric data to be used for forcing fields. Note that the WeatherFlow RAMS predictions are now all generated with equivalent or higher resolution WRF predictions.
Table 1. Current atmospheric models used in the MARACOOS weather prediction ensemble.
Institution / Model
Runs Used (UTC)
Rutgers University WRF
Sterling VA WFO WRF
WF RAMS Chesapeake Bay
WF RAMS Cape Cod
WF RAMS Atlantic Coast (New)
WF WRF Atlantic Coast (New)
WF WRF Atlantic Coast B (New)
The new information in this report relates to verification statistics performed on the ensemble and each of the individual members. The wind vector differences (WVD) were calculated on 6 groupings of stations within the MARACOOS domain including: Category 1- inland by 10 miles (10 sites), Category 2 – Inland by up to a mile (8 sites), Category 3 – At water’s edge over land (13 sites), Category 4 – At water’s edge over water (13 sites), Category 5a – Over water <1/2 mile from shore (11 sites, Category 5b – over open water (9 sites).
Table 2 contains wind vector differences for the 2015 year of model runs, including the AMAPS coupled ocean/atmosphere model developed by Dr. Rouying He at NC State University, and two versions of the Ensemble with and without the inclusion of AMAPS. In WeatherFlow’s experience, wind vector differences are sensitive to large directional differences, and a missed or miss-timed sea breeze event will penalize the prediction model. This explanation may be behind what otherwise would be some unexpected results. Sites in bold denote the best performance for either a geo-type (columns), or a model (rows). Conversely, sites in italics denote the poorest performance.
Table 2. Model Performance (wind vector difference) for 2015. Best performers are in bold text, worst in italics.
Model / Geo Type
AMAPS 9 km
Ensemble 2 km (no AMAPS)
Ensemble 2 km (with AMAPS)
HRRR 3 km
Nam 4 km
RAMS 2 km Chesapeake
RAMS 2 km NorthEast
WRF 2km LWX
WRF 3km Rutgers
WRF 5km ARW
WRF 5km CONUS
AMAPS (row 1), the only coupled ocean-atmosphere model, had significantly higher WVD differences across all geo-types. In fact, for the open ocean, the error approaches twice the WVD for the ensemble with AMAPS excluded. Most likely this can be attributed to the relatively low resolution of AMAPS and an inability to properly resolve the intricacies of the coastline. Despite its poor performance, adding AMAPS to the ensemble had little effect on the overall wind vector WVD difference (compare rows 2 and 3). Significantly all models perform their worst at the coastal boundary, performing best inland (compare columns Cat 1 with Cat 4), and marginally better further offshore (compare columns Cat 4 with Cat 5A/B). This confirms our long held opinion that it is the complex coastal boundary that is the most difficult to simulate.
B)HF-Radar Equipment: The Mid Atlantic High Frequency Radar Network performed well during this progress period. The uptimes of the individual sites in the network are provided in Table 3. The long-range average is on par with previous periods. The 13 MHz radar station at Sea Bright had the electrical installed and is operational.
The management of the stations in Delaware Bay (CMPT, HLPN and SLTR) was transferred from University of Delaware to Rutgers University on June 1, 2016. We are preparing to evaluate the location of the radars to see if the locations should change and will seek guidance from MARACOOS stakeholders.
The WHOI station on western Nantucket (NWTP) was installed in May 2016 and is now available on the National Network.
Table 3: Table of data availability for the progress period. The right column represents the % of data available on the National Network. All the funded operators maintained statistics on site visits that were necessitated by a station failure. There were 183 days in this progress period with 64 station visits. There were 38 stations managed (the WHOI sites were excluded). This computes to a mean time between failure of 109 days (183 days * 38 stations / 64 station visits). This is an improvement over the 67 days as reported by MARACOOS in the December 2014 progress report.
The spatial and temporal coverage of the 5 MHz long-range network is shown in Error: Reference source not found. The network again exceeded the 80/80 metric that was developed by MARACOOS and the US Coast Guard Office of Search and Rescue.
C) HF-Radar QA/QC: The MARACOOS HF radar operators have held twice monthly conference calls to exchange information and develop new procedures. The group maintains a status of site outages here:
Figure 1. Spatial and temporal coverage of the 5 MHz network from period December 1, 2015 to May 30, 2016. Teresa Updyke, Anthony Kirincich, Sara Haines and Hugh Roarty are serving on the QARTOD HF-Radar Quality Control Committee. The draft QARTOD HFR document was released on April 20, 2016, and a final version presented to NOAA/IOOS on May 24, 2016.
Figure 1: MARACOOS HF-Radar coverage for the progress period.
MARACOOS has started generating a quality controlled radial file. This process adds five columns to the ASCII file generated by the SeaSonde. The tests we are currently performing are the global range, trend, stuck sensor, and gradient test. We are also including a summary flag as dictated by the IOOS QARTOD document “Manual for the Use of Real-Time Oceanographic Data Quality Control Flags”.
Table 4: Lagrangian skill score for the seven drifters deployed in the Mid Atlantic in May 2016.
MARACOOS conducted a drifter experiment in cooperation with the US Coast Guard from May 10, 2016 to June 23, 2016. Seven drifters (four off Martha’s Vineyard and three off New Jersey) were released. The average skill score (Table 4) for the different data sources were tested using the modeled trajectory of the drifters after 48 hours. Those data sources are: 6 km product from the 5 MHz Network, 2 km product from the 13 MHz Network, 2 km product from the 13 MHz Network that is QCd, and the 9 km product from the HYCOM ocean model.
D) Underwater Gliders: The last six months have focused on planning and preparation for the upcoming summer and autumn IOOS glider deployments. Beyond the field campaign preparation, MARACOOS also conducted a regional glider deployment in the northern sector of the domain. Our report will highlight the planning efforts and results collected by the spring glider deployment.
Spring glider deployment. Many of the IOOS glider flightsin the Middle Atlantic Bight (MAB) fly 500 km triangles – as favored by the MARACOOS modelers. The UMass Dartmouth/SMAST/OCEANOL team deployed an ocean glider (Blue) just west of Martha’s Vineyard on 18 May 2016. For the following 3+ weeks, Blue sampled the ocean along a triangle in the Southern New England Bight (SNEB) region south of Martha’s Vineyard. Glider Blue’s mission was guided by the MARACOOS Glider Technical Center – in this case a primary collaboration between the staff at the Rutgers University (RU) central node and the University of Massachusetts Dartmouth (UMassD) secondary node. The glider Blue was outfitted to measure temperature (T), conductivity (C), oxygen (O), optical backscatter (OB), chlorophyll-a fluorescence, coastal dissolved organic matter (CDOM) and estimated section-averaged velocity (V). The glider successfully sampled the cold pool and nearshore coastal waters and the data is being compared to past years. This deployment serves as the baseline, before the major MARACOOS efforts in the summer and fall. The Blue glider was recovered June 10, 2016.
Planning for the fall and summer campaign. The MARACOOS glider team Rutgers (Kohut, Miles, Schofield), U Mass Dartmouth (Brown), U Delaware (Oliver), U Maryland (Boicourt), and Virginia Institute of Marine Science (Gong) have been planning the glider efforts for MARACOOS. Initial discussions have focused on providing coverage during the summer and fall, and spanning the full MARACOOS domain. This provides a large-scale mid-Atlantic regional context in which other funded glider efforts (see below) can be conducted. Currently there are four funded MARACOOS glider missions for the spring/summer/fall of 2016 (one FY16 funded and 3 FY17 funded). The spring mission was conducted out in the northern region of the MARACOOS domain. The summer missions will conduct zig-zag transects laterally across the shelf between Massachusetts and New Jersey, and the Chesapeake region, respectively. These along shelf zig-zag missions in the MARACOOS northern and southern sectors will be run in summer and fall. Among other objectives, the MARACOOS glider effort will be measuring the Mid-Atlantic Cold Pool. These data will be available in real-time via GTS to MARACOOS and other interested data assimilation modelers, who can map the variability in the Cold Pool during its 2016 season.
Several other glider programs will be conducted in in the mid-Atlantic in conjunction with the MARACOOS effort. The MARACOOS effort is being coordinated with these non-IOOS funded efforts, which have their own specific mission requirements. Through this coordination, the resources of the Mid-Atlantic community will be leveraged across these other programs as outlined below.
NJ DEP: New Jersey DEP will continue its series of summer coastal glider flights to survey the low oxygen nearshore waters. These missions, which will be conducted from July through October, provide valuable information on the inshore edge of the Mid-Atlantic Cold Pool that is measured by the MARACOOS effort.
NOAA CINAR: This is the final year of the current NOAA CINAR project that seeks to investigate the physics of large storms along the northeast United States. The NOAA CINAR gliders remain waiting for summer and/or fall storms that will impact the Mid-Atlantic and Gulf of Maine region. By late summer 2016, the CINAR gliders will be deployed to be ready for a late summer/fall storm or to complement the ongoing MARACOOS 2016 effort. CINAR team members are presently discussing alternate mission plans.
ONR-Biolum: This ONR project focuses on development of a new bioluminescence sensor. Discussions are currently focused on using a glider as the platform to carry the sensor, and there is a potential resource for MARACOOS glider pilots. This provides leverage to conduct four missions instead of just the three funded (FY17) missions. Current needs for this mission include a summer deployment when bioluminescence patterns are highest.
VIMS test flight: In Spring, there was a test-flight of a new Slocum Glider. The glider ran a cross-shore transect on the outer 1/3 of the continental shelf just off the southern Virginia coast.
All glider data collected by the MARACOOS team will meet QARTOD standards and contain full QA/QC information. Data will be distributed into the IOOS Glider Data Assembly Center.
E) Satellites: MARACOOS continues to support expanded satellite coverage from the Gulf of Mexico, to Cuba to Newfoundland. We utilize de-clouded sea surface temperatures (SST’s) and post the data to a publicly available THREDDS server in a Climate Forecast (CF) netCDF format. The MARACOOS CF-compliant SST data feed, which began in 2005, is updated in near-real time (http://tds.maracoos.org/thredds/SST.html). While the Rutgers University L-Band dish is currently offline, the University of Delaware dish continues to deliver data in near-real time. These data can now be accessed at greater speeds due to computer infrastructure updates on the THREDDS server. This real-time SST data is also visualized via the MARACOOS Asset Map (http://assets.maracoos.org), in Google Earth (http://modata.ceoe.udel.edu/web_kmzs/), and via browser-based map services http://www.ceoe.udel.edu/our-people/profiles/moliver/orb-lab/real-time-sea-surface-temperatures). MARACOOS is also deploying the JPL MUR SST product to our users. It is a gap-filled, 1km product but has a latency of 1-2 days. (http://tds.maracoos.org/thredds/SST.html?dataset=MURSST-Agg). The University of Delaware ocean color processing creates a real-time CF-compliant netCDF4 ocean color data feed, which is being updated in near-real time (http://tds.maracoos.org/thredds/MODIS.html). Ocean color data is now processed with NASA SeaDAS instead of the NRL APS system. The ocean color THREDDS feed includes 39 ocean color-related products, including estimates of the inherent optical properties, which are critical for understanding the coastal ocean. The ocean color record covers 2002-present. We have also fully operationalized a coastal salinity product based on ocean color data and it is available on MARACOOS THREDDS server.
Data streams from VIIRS and MODIS-Aqua are currently being delivered to Roffers Ocean Fishing Forecast Service (ROFFS) and Oceantemp.com. MARACOOS is these businesses’ primary data provider. The data processing stream has been ported to the University of Delaware Mills Super Cluster for satellite data processing. This has significantly increased the speed of data delivery. Data from the UDel Geo-Stationary satellite receiving dish is available on the web, and is now on an experimental THREDDS server hosted at UDel:
Rutgers has halted work on acquiring the 28-year AVHRR data set from the NOAA CLASS website due to a lack of resources. Rutgers is investigating leveraging funds to bring the X-band system back online and convert it to a dual L/X-band system that will acquire Aqua, Terra, NPP, JPSS, and all POES and Metop satellites. Rutgers continues to download the five channel level-1 data in order to enable reprocessing of SST data using a coldest pixel algorithm which eliminates the daily localized cold centers of east coast upwellings. Real-time acquisition and coldest pixel composites are continually being generated in near-real-time for use in atmospheric modeling.
F) Dynamic Models: Rutgers continues development of an expanded Gulf of Maine and MAB ROMS model to succeed the present real-time ESPreSSO system. The prototype system – termed “Doppio” – has been run in reanalysis mode for 2007-2016 to develop, test and refine the configuration of the data assimilation system to improve upon the ESPreSSO set-up. The strong tides and complex bathymetry of the Gulf of Maine and Georges Bank present a different dynamical regime from the MAB which has driven the necessity to develop a more robust configuration. As of May the system has been operating in real-time prediction mode. Various failures or difficulties with the data streams that inform the assimilation have been encountered - principally the intermittent unavailability of AVHRR SST and the occurrence of glider data being duplicated. Steps are being taken to harden the Doppio data pre-processing to these factors. Replacing ESPreSSO with DOPPIO as the ROMS MARACOOS real-time system will not occur until 6 months of stable operation, with skillful forecasts, is achieved.
The multiscale operational weekly predictions by the SMAST-HOPS model has continued with assimilation of ARGO and SST data during December 2015 through May 2016. In addition, in May 2016, data from Glider BLUE was assimilated, which continued in June. In a related effort, the past six years (2009-2015) of Gulf Stream north wall (GSNW) and eddy locations were used to develop an automated method of identifying the GSNW and Rings using altimetry data, which will be used for situations when SST data might be unavailable due to clouds. Finally, a new version of MIT-MSEAS model (modified version of the parent code of SMAST-HOPS) is being ported to SMAST for operational usage.
As a prelude to the 2016-2018 efforts, a WCR (warm core ring) census effort has been started while initiating the metadata creation of the 40 years of Gulf Stream charts donated by Jenifer Clark to the Ocean Modeling and Analysis Laboratory at UMass Dartmouth. The ring census will document information related to each WCRs’ birth, progression path, size and absorption/dissipation. Each ring is named using a unique combination of alpha-numeric characters, WEyyyymmddX, where WE stands for Warm Eddy, yyyy is the year of birth, mm is the month of that year, dd is the day of the birth-month and X is an alphabetic character assigned to the Ring in its chronological formation during that year. Table 5 presents an excerpt from the recently completed WCR census for 2015. In 2015 there were more than 23 WCRs seen in the first seven months. During July a maximum of seven new WCRs were formed. We do not know whether this is a typical pattern or an anomaly driven by higher temperature over spring/summer of 2015. Another interesting feature is the seven-month-long presence of a single WCR near the shelf (Ring B), which might have impacted the cross-shelf exchange over at least three seasons (winter, spring and summer).
A complete WCR census such as above when available for 2000-2015, will be extremely valuable for understanding WCR impact on the MAB shelf, slope, GOM, and GB regions. This ring census will also help advance the predictive capability of all MARACOOS and other IOOS modeling efforts in this region.
In May 2016, work on the MARACOOS-leveraged Stevens NYHOPS estuarine model – one of the 125 operational members of the Stevens Flood Advisory System (www.stevens.edu/SFAS) that produces 78hr ensemble-based forecasts every 6 hours based on various hydrometeorological forcing components – was completed. The member then became part of the operational ensemble suite and became available over an open-access THREDDS server:
Table 5. Ring activities during the first four months of the year 2015 are shown below. DOB = Date of Birth or formation; DOA = Date of Absorption or dissipation.
The MARACOOS supported NYHOPS ensemble member uses meteorological and coastal boundary forcing from two MARACOOS-related research products: The Rutgers WRF meteorological model and the Rutgers ESPRESSO coastal ocean model. Out of the 125 members of the Stevens flooding ensemble, and over a 4 month evaluation period preceding operational deployment of the MARACOOS-leveraged member, that member had the third best root-mean-square-error in storm surge prediction among all deterministic members considered, closely following in skill the NAM- and ECMWF-High Resolution- forced members. Results of the Stevens Flood Ensemble research have produced many accepted publications in major journals, with many more in the submittal stage. NYHOPS was also accepted as the NSF COOPS OFS for NY Harbor and the NWS source on their public website.
In June 2016, an open-access, multi-decadal hindcast of daily hydrodynamic properties within Mid-Atlantic estuaries will be completed using the NYHOPS 3D hydrodynamic model. Both operational predictions, and hindcast model data and associated climatologies and anomalies are available online at http://colossus.dl.stevens-tech.edu/thredds/catalog.html, and are in high demand for navigation, flood forecasting, regional climate change and teleconnections research, fishery Habitat Suitability Studies, adaptive fishery surveys, ecosystem modeling, water quality modeling, and engineering planning.
G) DMAC: The WAF of MARACOOS services continues to be registered with NGDC and is listed in the IOOS catalog. The MARACOOS services are now also registered with the new IOOS CKAN Catalog at data.ioos.us. In total MARACOS has 924 services including: 468 DAP, 427 SOS & 29 WMS.
The TDS server (http://tds.maracoos.org/thredds/catalog.html), which is currently running version 4.6.4, includes the following data: UDEL SST (NOAA AVHRR, MODIS, and MURSST), UDEL Chlorophyll (MODIS), UDEL Regional Salinity Products (MODIS – Chesapeake Bay, Delaware Bay, Mid-Atlantic), WeatherFlow met model data (ensemble, RAMS Chesapeake Bay, RAMS North East, NAM4k, and HRRR), and Rutgers WRF met model results (3km and 9km). SciWMS services are available at wms.maracoos.org for AVHRR, MODIS, and MURSST and also for the MARACOOS Espresso product hosted at Rutgers. The TDS server is monitored and data managers receive twice daily reports of any dataset older than 30 hours as well as notifications if services go down. In spring 2016 the TDS was reorganized to improve performance and availability. Previously, indexing of the Rutgers WRF GRIB files was a large source of memory drain. These GRIB files are now stored in individually-dated directories, which in conjunction with updating the TDS to version 4.6.4 allowing the GRIB indexing to occur on a separate machine using the THREDDS Data Manager (TDM), limits the memory usage. The Rutgers WRF archive now includes data back to May 2013 without taxing the system. The satellite data has also been reorganized into real-time and archive directories. The real-time directory stores the last 8 days of data. The archive directory stores older data in individual yearly directories.
The DMAC team continues to support a SOS server at http://sos.maracoos.org/stable/catalog.html. This SOS server provides data via ncSOS services using ncSOS v1.1. The SOS services deliver data from 9 WeatherFlow wind stations, 10 Hudson River Environmental Conditions Observing System (HRECOS) stations, and dozens of USCG surface drifters (SLDMBs).
A beta version of MARACOOS OceansMap was released in May 2016 (oceansmap.maracoos.org). OceansMap is MARACOOS’s flagship platform for collecting, analyzing, and disseminating met-ocean information. By collecting scientific data in disparate formats and making it available via standard web services, OceansMap removes hurdles for finding and accessing metocean data. It provides access to a variety of environmental datasets including in situ and remotely sensed observations, glider profiles, and global and regional model predictions. OceansMap allows users to integrate both observation and model data, to set time frames and areas of interest, visualize all available data, including both observation profile and 3D model data, create interactive time-series and profile plots, and perform model-observation comparisons. The DMAC team has also developed a new Python tile server and has started serving the AVHRR data as tiles, which provides greatly improved performance and visualization.
The GliderDAC 2.0 has continued to progress in 2016. A new home page for the DAC is now available at gliders.ioos.us and the glider network map has been ported to the Oceansmap framework and is available at gliders.ioos.us/map. The DAC team coordinated with NAVO to ensure that Navy glider data are displayed on the glider map. The team also began planning for QARTOD implementation including flag variable storage, defining tests appropriate for application by the DAC, and coordination of test implementation with the Rutgers glider team. Operations and maintenance tasks for GliderDAC included removal of the requirement to define the WMO ID when setting up a deployment, adding attribution to glider profiles in the glider map, and improved the rendering of the glider profile cross-section plots.
H) Education and Outreach: MARACOOS Data Stories: In the coming reporting period the MARACOOS education team will propose the development of 5 data stories that utilize MARACOOS data for a range of stakeholder audiences. This work was delayed as sufficient resources were not available during the last reporting period. These stories will be loosely modeled after the data stories developed for the Polar ICE program (see http://polar-ice.org/focus-areas/polar-data-stories/what-drives-patterns-in-ocean-change/). The intended audience for this polar data story was science-interested members of the general public that come to the website.
Stories will be developed to help make MARACOOS information and data impactful, memorable, and personal (i.e. help particular stakeholder groups intellectually and emotionally connect to the mission of the MARACOOS program). The intended audience for the data stories is interested members of the public, MARACOOS stakeholders including MARACOOS scientists, policy makers, and regional coastal managers. The goals and intended audiences will vary across the data stories.
We are aiming for the user to spend about 3-5 minutes interacting online with the data story. The stories will be delivered primarily through interactive data visualizations that provide guidance to the user to understand the visualization and support their exploration of the data as well as to provide opportunities for more open-ended data exploration. The stories will be developed with the Martini Glass or Interactive Slideshow approach. They will have one or more of the following design themes: Narrate Change Over Time, Highlight Contrasts, and Connection to the real-time MARACOOS portal. The five stories are focused on rip currents, storm intensity/inundation, fisheries management, water quality and offshore wind harvesting.
I) Economic Benefits:MARITIME SAFETY: Navy Fleet Weather Center Norfolk (FWC-N): The MARACOOS Asset map continues to be used on the watch floor of FWC-N. Our challenge as a supporting observing/modeling research community is to help Navy determine go/no-go sortie decisions 72 hours before the onset of 50 kt sustained winds and 2-4 foot storm surges. This is particularly important during the east coast hurricane season.
NOAA COOPS PORTS: MARACOOS continues to work with NOAA COOPS to include surface current data into NOAA PORTS as a product to serve the maritime pilots and navigation community in New York / New Jersey Harbor and the Chesapeake Bay approaches. Although the specific economic benefit is difficult to assess, it is reasonable to assume a risk reduction and complementary advantage to mariners in their efficiency, reduction of loss, and bottom line operating expenses, resulting in an economic advantage to the transportation and commerce sectors operating in those port areas.
ECOLOGICAL DECISION SUPPORT: In December 2015, MARACOOS joined the inaugural Atlantic Mackerel Working Group meeting in Point Judith, RI. This meeting was an industry-science-resource manager collaborative focused on preparation for the 2017 Atlantic mackerel federal stock assessment. This Mackerel meeting built on the foundation and trust relationships established with the Butterfish-Squid Seascape project funded by NMFS Cooperative Research, Industry (GSSA) and MARACOOS. As a result, a collaborative team was formed and a proposal was developed to the MAFMC Collaborative Fisheries Research RFP, titled: Collaborative development of a winter habitat model for Atlantic Mackerel, “version 2.0”, for the identification of “cryptic” habitats and estimation of population availability to assessment surveys and the fishery. The project was awarded funding in April 2016 and is designed to provide an ecosystem perspective to the 2017 Federal Atlantic mackerel stock assessment, with the potential of real economic benefits. The work will rely on IOOS and MARACOOS assets. The team is composed of PIs drawn from industry, science and management.
COASTAL INUNDATION: NOAA NOS: MARACOOS has been cooperating with Old Dominion University and Virginia Institute of Marine Science for MARACOOS to serve as the data integrating agent in support of the Hampton Roads Sea Level Rise Intergovernmental Pilot Project. The projected economic benefits associated with a reduction in loss can be partially attributed to the data collection and integration efforts of MARACOOS. As funding sources are identified, collaboration discussions continue, with strong involvement from NOAA NOS, especially COOPS, and involvement from Sea Grant, State and local agencies, and other federal partners.
NOAA NOS: Sustained exchanges with Stevens Inst of Technology and the National Ocean Service over the last year have resulted in a Memorandum of Understanding with NOS that officially accepts the NYHOPS model, developed and being run operationally at Stevens Institute of Technology, as part of the NOS OFS model guidance suite (e.g. http://www.weather.gov/phi/marine; click “New York Harbor”). This work will further enhance NOAA products and the economic benefits they provide to the maritime transportation sector, as it conserves and leverages federal resources.
MATOS: Please see the current Supplemental report for more information on MATOS.
J) CBIBS Buoys: The Chesapeake Bay Interpretive Buoy System is being used as a development tool for MARACOOS support of fixed platform time series data, including Web Services (SOS), QARTOD QA/QC recommendations, and provision of data to Federal agencies. The CBIBS Data Acquisition software is now running on the Applied Science Associates cloud; and a new CBIBS database with Web Services has been constructed, populated with historical data, and is being updated in real-time. NOAA, MARACOOS, and ASA are jointly participating in IOOS DMAC QARTOD conference calls, and can be viewed as a lead RA in implementing QARTOD standards. A description of the MARACOOS CBIBS Data Management and QC system was presented at the IOOS DMAC Workshop in May 2016. Data are served via IOOS-standard SOS web services, and QARTOD QA/QC is being applied in real-time. QARTOD QC parameters are being fine-tuned for each buoy and measurement type. Through MARACOOS, ASA has been tasked with developing a complete end-to-end system for collecting, quality checking, archiving, and delivering CBIBS data, as well as documenting the system for IOOS certification purposes. CBIBS Buoy data now appear in full on the MARACOOS Asset Map directly from the QC’d database (as opposed to coming through NDBC), making access more timely, accurate, and complete. Discussions proceed with NCEI on ways to archive QARTOD flags, again leading RA activities in this field.
ASA supported the NOAA Chesapeake Bay Office in transferring the CBIBS system to NOAA servers. In addition to CBIBS physical data, the database supports collection of Acoustic Tag data from the CBIBS system in support of the Mid-Atlantic Acoustic Tagging Observation System (MATOS), a component of the IOOS ATN.
MARACOOS continues to provide support for CBIBS long-term planning and operations and maintenance, as well as expanding operational and prediction activities in the Chesapeake Bay. Discussions have been held with Chesapeake Bay Program, shellfish user groups, CB Sentinel Site Cooperative, and ecological forecasters to integrate CBIBS data into their activities and to look at new observing assets and tactics to improve Chesapeake Bay management products. NOS CO-OPS routinely uses CBIBS real-time surface salinity observations to correct the CBOFS hydrodynamic model used for Chesapeake Bay PORTS. Additions to standard CBIBS data collection will test CBIBS data input into the NOS PORTS system, with an ion-water test expected in Summer 2016. In collaboration with the University of Delaware PI Wei-Jun Cai, a SeaPhox pH sensor was been installed on the bottom mooring at Gooses Reef. Preparations are underway (planning, in-kind equipment provision) for a MARACOOS-CBIBS Ocean Acidification buoy near the mouth of the Chesapeake Bay. Testing has begun on instrumentation for supporting the proposed 2017 MARACOOS RTAP Hypoxia Forecast Model project.
K) QARTOD: MARACOOS continues to work toward implementation of QARTOD manuals as a replacement and/or enhancement of existing QA/QC processes, including in support of MARACOOS' certification application.
QARTOD recommended test and associated flags is now implemented on the real-time MARACOOS HF-Radar radial data streams. The file format with the quality flags has been sent to the HF-Radar National Network for their review and approval. QARTOD recommended quality flags are also being implemented for the CTD and DO glider data. We have an approved NetCDF format that will be delivered to the National glider DAC beginning this summer. For more information on HF-RADAR QARTOD work, please see the HF-RADAR section of this document.
For CBIBS QARTOD work, please see section J) CBIBS in this document.
SCOPE OF FUTURE WORK
The MARACOOS no cost extension, submitted in April 2016 and approved in June 2016, details out the scope of the continuing work on the project through May 2017.
PERSONNEL AND ORGANIZATIONAL STRUCTURE
The total Board complement is currently at 14 with the potential for up to 4 additional board-appointed members in the future. During this period, Board Members Paul Cooper, Michael Bruno, and Jay Odell retired from the Board. Mary Yates was hired during the reporting period to start on June 1, 2016 as MARACOOS Program Coordinator, while 2 interns (Katie Liming, a UDel Junior, and Jessica Jenkins, a UDel Freshman) carried out social media and information research activities.
The total budget for years 1 through 5 is $13,935,022. Year 5 subcontract totals for the 16 subcontracts and Rutgers, the prime, are shown in Table 6. Current balance remaining for the lead and all subcontracts as of June 2016, is shown in the right hand column. Note that many of the subcontractors billing lags several months, so these totals are not indicative of the total spent to date.
Table 6. MARACOOS years 1-5 budget distribution with total and remaining budget listed for the prime and all subcontractors.