Rutgers University
  • Navy Gliders View the Cold Pool

    Posted on October 25th, 2019 Mike Crowley No comments

    Two Navy gliders were deployed in the Middle Atlantic Bight (MAB) between September 11 & 26, 2019. Their cross-shore trajectories were planned to capture the cold pool, a cold bottom layer of water that characterizes the MAB at this time of the year. There transects are shown below.

    In the figures below, the top figure shows the cross-shore temperature transect for glider ng618. In this figure the cold pool can be identified as the bottom water with temperature less than 10 degrees Celsius. The cold pool has important implications for the fisheries of the region and it can become a “fire extinguisher” for tropical cyclones by lowering the sea surface temperature through vertical mixing during the passage of a storm.

    The bottom figure shows the same ng618 glider trajectory from the Global Ocean Forecasting System (GOFS 3.1) that is a global operational ocean model run by the US Navy. The comparison between glider observations and the GOFS model are remarkably good.

  • Low Surface Salinity During Tropical Storm Karen

    Posted on October 3rd, 2019 Mike Crowley No comments

    During Tropical Storm Karen, there were 10 gliders deployed in the Caribbean (below). On Sep 24, the eye Karen passed over Navy glider NG278, just south of St. Croix.

    There is a good agreement in the vertical structure of temperature between GOFS 3.1 and the glider temperature (below).

    The GOFS 3.1 model, however, has a surface fresh layer (barrier layer) that is not as sharp as the observations show. This barrier layer gets even fresher and shallower during Karen (below). At the moment we are investigating if this freshening can be attributed to increased precipitation during the storm. If this is the case, this can be a mechanism that reinforces stratification during a storm and inhibits vertical mixing. We would like to investigate how important this possible mechanism is in controlling the storm intensity.

    The freshening that happened during Karen (Sep 23 – Sep 24) is not initially captured by any of the models (below).

    By Sep 26, the surface salinity in GOFS and Copernicus were much closer to the observations (both models assimilate data) but RTOFS still lags behind (below).

    Images of these daily comparisons, since the inception of the 2019 hurricane season, are available here.

  • There are a lot of gliders out there!

    Posted on October 1st, 2019 Mike Crowley No comments

    The 2019 hurricane season is in full swing in the Atlantic Ocean. Since the inception of the hurricane season on June 1, 42 ocean gliders have been deployed in the western Atlantic ocean by numerous academic, corporate and government partners. These gliders have acquired over 68,000 profiles of temperature and salinity to date. Some gliders also record data on water optics, oxygen, current speeds, pH, and wave heights. A distribution of the groups funding the deployments is shown below as well as a map of their locations.

  • Salinity Adjustment Ahead of Hurricane Dorian

    Posted on September 27th, 2019 Mike Crowley No comments

    Accurate ocean initial conditions in numerical models are necessary to improve hurricane intensity forecasts (see article). One way to improve the initial conditions in operational ocean models is to use data assimilation, a technique that allows a numerical model to ingest observational data to nudge the model predictions closer to reality. The goal of the comparisons is to evaluate how data assimilation is impacting the model performance and ultimately how ocean processes affect storm intensity. Images of these daily comparisons, since the inception of the 2019 hurricane season, are available here.

    For the daily comparisons we use three operational ocean models: Global Ocean Forecasting System (GOFS 3.1, Navy), Real Time Ocean Forecasting System (RTOFS, NOAA) and the operational Mercator global ocean analysis and forecast system (Copernicus).

    During Hurricane Dorian there were a total of seven gliders deployed in the Caribbean reporting to the IOOS glider data assembly center (IOOS Glider DAC). Once the gliders start reporting data to the IOOS glider DAC, the data is sent to the Global Telecommunication System (GTS) where it can be assimilated by ocean models.

    On August 29, Tropical Storm Dorian passed over three gliders: SG665, SG666 and SG668 (Figure 1).

    Figures 2a-2d below highlight the comparisons of the models to glider SG665. In general, the operational ocean models capture the temperature structure in the Caribbean (Fig. 2a and 2b), however, in this example, the models tend to underestimate salinity at the surface and subsurface salinity maximum around 200 meters in this area. The assimilation of glider data into operational ocean models plays a very important role in correcting the subsurface and surface salinity structure. An example of this correction is highlighted in figures c and d. GOFS 3.1 (red line) is shown in figure c on the first full day of the glider deployment. Figure d shows a shift of 0.4 psu after a week of assimilation, pushing GOFS 3.1 much closer to the reality of the salinity structure.

  • International Challenger RU29 Mission Update

    Posted on September 12th, 2019 Mike Crowley No comments

    RU29 is tuning up to measure currents: RU29 spent the day profiling in about 1000 m depths off Virgin Gorda, with operators tweaking settings on the onboard Doppler current velocity profiler to optimize its ability to measure ocean currents. An important aspect of the mission will be measuring speed and direction of currents below the surface to determine the movement of the water masses distinguished by their temperature and salinity characteristics.

    Currents in the region can be strong and have complex patterns. The figure below shows the modeled (US Navy AMSEAS model) currents in the region for 11 September, 2019. RU29 is in an area of northward surface currents greater than 1 knot. Ocean gliders themselves are only capable of forward speeds of around 0.5 knots, and only when they are ascending or descending. In order to make headway, they depend on spending most of their time well below the surface, where the currents are generally slower. If necessary, the glider does have a small electric motor and propeller which it can use, but at the cost of tapping into its limited battery reserves.

    RU29 recent trackline and position on 11 September 2019. The glider should begin its trip across the Anegada Passage station line on 12 September.

    Modeled surface current vectors from the US Navy AMSEAS model from 2100 UTC, 11 September, 2019. From OCEANSMAP Caribbean.

    To see all of the current reports, you can view them here:

    The International Challenger RU29 Mission is a collaboration among Rutgers University (New Brunswick, NJ) Center for Ocean Observing Leadership, Ocean and Coastal Observing – Virgin Islands (OCOVI, an affiliate of the US Integrated Ocean Observing System (IOOS) Caribbean Coastal and Ocean Observing System (CARICOOS)), and the University of the Virgin Islands. This mission will collect upper ocean data to improve hurricane forecasting models; study conditions in the Anegada Passage, an important region for NE Caribbean climate change; and build international cooperation for ocean observing and glider activities. The mission will work in the waters of the US Virgin Islands, British Virgin Islands, and Anguilla, under international Marine Science Research permit . Funding for the project comes from the U.S. National Oceanic and Atmosheric Adminstration through the IOOS, Rutgers University, and the G. Unger Vetlesen Foundation.

  • Ocean Response to Winter Storm

    Posted on November 16th, 2018 Hugh Roarty No comments

    A storm moved through the Mid Atlantic on November 15, 2018.  The MARACOOS High Frequency radar network captured  the center of the low pressure system as it moved north past Delaware and New Jersey.


    The surface winds from the Global Forecast System (GFS) had the center of the low further inshore.


    We’ll continue to analyze these two data sets to understand the discrepancy.

  • Rapid near surface temperature change in Michael

    Posted on October 18th, 2018 Scott Glenn No comments

    Here is a comparison of the temperature in the upper mixed layer (10 m depth and 20 m depth) at the location of Navy Glider 288.  Hurricane Michael passed within less than 50 km of ng288.  The swath of tropical wind scale winds was about 500 km wide, so this was relatively close to the track of the hurricane.  And the glider was on the right had side of the track, so perfectly placed in the most intense winds.

    Solid black vertical line is time of hurricane eye passage over Navy Glider 288. Dashed black vertical lines are data assimilation times.

    The agreement between the surface layer temperature observed by ng288 and the GOFS 3.1 forecast is remarkable. The model even appears to be cooling the surface layer at the proper rate.  Note that most of the observed cooling occurs after eye passage, the opposite of what we see in the Mid Atlantic.

    The dashed line at 1200 on Oct 10 will be an interesting case to study with our colleagues. We assume the glider data is being assimilated, our friends at NRL are good at checking for this.  During this time of rapid temperature change, if glider data from earlier in the day is assumed to apply at noon on Oct 10, it will try to increase the temperature as we observe here.  The model then appears to cool the temperature at the right rate despite being incorrectly offset boy about 0.5C.

    Establishing what happened through the data assimilation during the passage of Michael over ng288 will be a great research question for us to prusue over the winter as we prepare for the 2019 hurricane season.



  • Time series of near surface temperature from Navy Glider

    Posted on October 12th, 2018 Scott Glenn No comments

    Here we take a time series look at the near surface (2o m depth) temperature reported by the Navy Glider as Hurricane Matthew went by.

    The time of eye passage is shown by the vertical dashed line. Temperature increase is at one slower rate until about 05:00 on Oct 10, just before eye passage.  We then see a more rapid temperature decrease for about 8 hours, most of which occurs after eye passage.  By about 17:00 the near surface temperature starts to increase slowly.  So three ocean response time periods.


    Next to check the model ….

  • Global Model profiles for Hurricane Michael

    Posted on October 12th, 2018 Scott Glenn No comments

    Below is the same plot as the Glider CTD data but for the Global Ocean Forecast System (GOFS) 3.1 model.  Remarkably similar ocean.  Differences are in the details.


    The 26C isotherm here is at a depth of about 90 m.  The Navy glider has it at about 110 m, a bit deeper.

    Surface layer cooling occurs more slowly in the model.

    We will be looking into those details next.



  • Detailed CTD profiles in Hurricane Michael

    Posted on October 12th, 2018 Scott Glenn No comments

    Maria is looking back at the CTD profiles collected during Hurricane Michael as it approached the Gulf Coast.  In the figure below we have 2 days of CTD data from October 9 and 10. The 26C isotherm is indicated with a black line. The vertical black line 0n October 10 at 06:00 is the time of eye passage over the glider. The isotherms deepen most rapidly right after eye passage until about 09:00. Surface temperatures over the 2 days start between 29-30C and end between 27-28C.



    We will look at the time series of temperature at a depth near the surface, like 20 m, soon and get it posted.