Rutgers University
  • Hurricane Iota

    Posted on November 16th, 2020 Scott Glenn No comments

    Just a quick placeholder post as we approach the winter analysis phase.  Yesterday’s forecast for Iota intensification looks to be happening today. Looks like the hurricane forecasters are making good use of the HWRF guidance. Their discussion was about  high sea surface temperatures, low wind shear, and high moisture contents.

  • Hurricane Eta – Low Wind Shear, High SST

    Posted on November 2nd, 2020 Scott Glenn No comments

    Today’s 10 am Discussion on the National Hurricane Center reflects a common theme:

    "Low vertical wind shear and warm sea surface temperatures ahead of Eta are expected to 
    allow for continued rapid strengthening during the next 12 to 24 hours."

    We see  the forecasters are keying in on the relative impacts of vertical wind shear and sea surface temperature to help them turn the intensity guidance they get from decision aids into their intensity forecasts.  Current shear and SST conditions are prompting the call for a Cat 4 hurricane:

    "The updated NHC intensity forecast is above all of the 
    intensity aids and now calls for Eta to become a category 4 
    hurricane before it nears the coast of Nicaragua."

    Step 2, lets look at those conditions in OceansMap.  We see a rich mesoscale across the central Caribbean as Eta tracks from the Windward Islands to Nicaragua. Even the green colors in this image are above 28.75C.

    Step 3, check out the regional hurricane model forecast guidance:

    Here is the most recent posting on the website for HWRF and HMON (intensity on upper right), with HMON a low Cat 3 and HWRF a high Cat 3. The short black line indicates that the best track is following the growth curves for HMON and HWRF.  Official forecast in this forecast interval is taking Eta up to the edge of Cat 3.

  • Tropical Storm Zeta

    Posted on October 25th, 2020 Scott Glenn No comments

    The 11 am National Hurricane Center discussion notes the importance of the ocean impacts on Zeta’s intensity in three locations.

    1. In the current location, the low shear and high ocean heat content in the western Caribbean favor intensification.
    2. As Zeta crosses the central Gulf, increasing shear and diminishing ocean heat content is expected to prevent intensification.
    3. As Zeta approaches the northern Gulf, the unfavorable ocean and atmospheric conditions could cause zeta to weaken beyond what is in the official forecast.

    Below is the present regional guidance (HWRF in purple, HMON in green), and what we anticipate will evolve into future guidance, the experimental HAFS model (light blue). The tracks (left image) from these three models have converged since last night, and that makes them easier to compare.  Zeta should be going over similar oceans in these models, something to check.  On the top right is the intensity plot. The official intensity forecast (red) is above the model consensus. HAFS looks to be below the official forecast, keeping Zeta below hurricane strength as it approaches Yucatan.  The operational models are showing the Zeta intensifying as it approaches Yucatan, with HMON going to Cat 3 and HWRF going to a high Cat 2.   These models appear to be the ones that include the impact of the high ocean heat content in the western Gulf that the forecasters are saying favor intensification.


  • Tropical Depression Twenty-Eight

    Posted on October 24th, 2020 Scott Glenn No comments

    The 5 pm NHC forecast discussion reminds us that as this storm approaches the northern Gulf, the cooler shelf waters and the increasing wind shear will likely weaken this storm.

    Lets take a look at those cooler temperatures by checking the Navy’s GOFS model SST.  Here we see the official forecast track heading through the Yucatan Straits, with the cooler water way up on the shallow shelf of the northern Gulf.  The green colors are still up around 27C, so the darker blues below 26c are really up by the coast.


    Now lets look at what the forecast guidance is telling us.  HWRF and HMON, the two operational regional models, are hugging the right side of the cone, hitting western Cuba and tracking closer to the very warm (near 29C) waters of the Loop Current.  Both of these models go to hurricane strength, with HMON hitting Cat 2 for a short period while the tracks are still very close (science question for the winter). If these tracks hold the storm will pass over NDBC’s glider SG601.  The experimental HAFS model hugs the left side of the cone, tracking over Yucatan, and the cooler Mid-Gulf waters, remaining below hurricane strength throughout.


  • Five Gliders in a Picket Line

    Posted on October 8th, 2020 Scott Glenn No comments

    First a quick view of the northern Gulf the night before landfall of what looks to be a major hurricane. The image below shows the Sea Surface Temperature forecast for the Navy’s data assimilative Global Ocean Forecast System, a consistently high performing ocean model,  definitely a good choice for the start of our U.S. hurricane forecast workflow. It nicely demonstrates the value of the hurricane glider picket line concept.  These gliders have been deployed for weeks, feeding their data into the GOFS model everyday to do everything we can to put the best representation of the ocean we can underneath our best hurricane models.  We have two Texas A&M Gliders on the west, two Navy gliders in the middle, one NOAA glider piloted by USM on the east.  Multiple government, university and industry groups working together for the common purpose of generated the best forecasts possible. We call it “Distributed Autonomy”.  With sampling distributed in space and time.  With operations distributed between multiple shore labs. And sponsorhips distributed between different agencies. What we see below, the pre-storm conditions established by the GOFS model, is one reason why we deploy the Hurricane Glider Picket Line each year – to improve the hurricane models that we are using today to forecast this very storm.



    The second reason we deploy the glider picket line each year is shown below. It shows tomorrow’s GOFS forecast of the ocean currents, with the eye of Delta clearly visible a bit to the west of the official forecast, and the ocean current response to the downwelling-favorable winds (remember that blog entry from earlier in the week on upwelling versus downwelling?) along the LA-TX shelf, and we see the size of the storm.  Four of the gliders will be directly in the thick of it, gathering the science data that will help us improve the hurricane models of the future.  All that data to improve those future models is going to be gathered in the next 24 hours. By robots.  No people have to go to sea.  We are not overburdening the hurricane hunters and their flight crews. But we are there.



    Stay safe Louisiana.


  • When Guidance and Forecasts Align

    Posted on October 8th, 2020 Scott Glenn No comments

    Another morning with remarkable agreement between the two operational regional models (HWRF and HMON) and the official NHC forecasts (red). Both models are skimming the bottom of the Cat 4 wind speed, and the official forecast is putting it in the middle of Cat 3.  Not sure why the experimental HAFS model does not have the right initial condition.  Something to check over the winter.


  • Hurricane Delta and the Barrier Layers

    Posted on October 7th, 2020 Scott Glenn No comments

    With multiple gliders in place, lets take a look at how the global models are doing with the data.  We’ll focus on the two Navy gliders on the right hand side of the forecast track.  NG645 is more south, almost central Gulf, and ng314 is further north, still in deepwater, but closer to any freshwater sources.

    First the two temperature profiles.  First is glider 645, near perfect agreement between the glider profiles in blue and the global ocean models – GOFS, RTOFS and Copernicus.  For glider 314, we see a small difference between the temperature profiles in the 200m – 400m depth range.  The data and the European Copernicus keep the warmer water deeper than the two US models.  The upper ocean in all three look the same.



    Now we check the salinity for the barrier layers.  The southern glider 645 has lower surface salinity than at depth, but not a strong barrier layer like the European Copernicus.  Both US models are close to the glider data at ng645.  But look what happens to the salinity field closer to the freshwater at ng314.  Here the Navy glider says there is a fresh barrier layer, and the Copernicus data assimilation is doing a good job of following this glider profile, both in the barrier layer and below.  The US models are missing the barrier layer at 314, and are having trouble matching the salinity throughout the upper 400 m.

    The conclusion for our winter lookback is to look more closely at the salinity structure.  The previous blog posted noted the barrier layers based on the models were close to shore.  Here Navy glider 314 says the barrier layer extends farther offshore than the models are showing us.





  • Hurricane Delta in the Regional Forecast Models

    Posted on October 7th, 2020 Scott Glenn No comments

    Remarkable agreement this morning between the two operational hurricane forecast models HWRF and HMON.  Both have similar tracks, something we see often, and this morning both are also agreeing on intensity plotted in the upper right in the figure below.  Both HWRF and HMON (purple and green lines) have Hurricane Delta just below Category 4 as it crosses the warm waters of the Gulf.  The 4 am CDT NHC Discussion notes that they are following the guidance from the stronger regional models, and we see that in the red line for the official forecast.  The Discussion credits the high ocean heat content and the low upper level wind shear resulting in favorable conditions for an intense storm.  Both models also rapidly drop in intensity after moving inland, but what happens just before landfall? The NHC Discussion highlights the importance of the competing ocean and upper level shear processes to decide if there will be any weakening just before landfall, noting there there is an increase in the upper level shear, and there is slightly less heat content in the ocean as the hurricane approaches the coast. The new experimental HAFS model, the light blue line below, has a more westward track and a significantly different intensity – lets mark that as a good discussion point for the winter reanalysis of this hurricane.

    Below is the OceansMap view of that Sea Surface Temperature distribution.  Delta if forecast to track right up through the center of the warmest surface temperatures in the western Gulf.  It then crosses the slightly cooler waters approaching landfall, but even the blue shade below is above 27C.  We do not see surface temperatures below 26C until right up by the coast.  Also looks like we have 5 gliders nicely aligned cross track for the Hurricane Glider Picket Line, with 3 gliders spanning the track uncertainty cone, 2 gliders further off in the high wind areas to the left and right.  The NHC Discussion notes that they expect Delta to grow in size as it approaches the northern Gulf.


    Next we look at the Sea Surface Salinity from the Navy GOFS Model.  Look for any low salinity barrier layers that inhibit mixing and contribute to intensification.  We see that the surface salinity is relative uniform across the Gulf with the low salinity barrier layers confined to the areas along the northern Gulf coast.  So the band of slightly cooler surface waters that could promote weakening are also covered with a low salinity barrier layer that could promote intensification.  Another test case for the winter reanalysis.

    Now we look at the subsurface temperatures along track in the global ocean models where the ocean data assimilation takes place. The automated processing follows the hurricane track and plots the upper 350 m of the ocean temper field along the track.  White areas are when the hurricane is over land. First is the Navy GOFS model that does the actual assimilation. The NOAA RTOFS model (second plot)  looks back 2 days in the Navy model and then moves forward without assimilation but with the NOAA winds layered in.  The global RTOFS then provides the initial condition for the coupled regional ocean model in HWRF and HMON.  The first thing we note is that there is very little difference between the two models along the storm track – a good thing.  The gliders and argo floats have all been deployed well in advance, and there has been plenty of time for the data to work its way into the assimilative model as well as the downstream NOAA forecast models.  The two figures below also plot the location of the 26C isotherm as the black line.  The depth of this isotherm used to calculate heat content is relatively constant across the Gulf, hovering around a depth of 50m-60m.  Temperatures above the 26 C vary as you move north along the track but are relatively constant in the vertical in this upper layer down to this 26C isotherm.   So SST is going to be a good proxy for the heat content variations pre-storm. We also see that only as the storm is crossing the shallow continental shelf waters nearshore does the temperature drop below 26C.  Watching how this water reacts as the storm approaches is one thing we will be watching.  As Delta approaches the coast from the south, these coastal waters will experience strong winds to the west. Alongshore winds to the west are downwelling favorable.  Downwelling can move this cooler nearshore water off the shelf along the bottom, and replace it with warmer water transported in from offshore.  Thus the hope that this band of cooler water may offer some protection assumes that the water is stationary, that it does not move.  With strong nearly Cat 4 alongshore downwelling favorable winds from the leading edge of Delta, that water is going to move. Depending on the forward speed of Delta at the time, that water may have plenty of time to move well ahead of the eye passage. Just like we saw with the glider in Hurricane Sandy.





  • Sally tracks northwest, SG601 flies south

    Posted on September 14th, 2020 Scott Glenn No comments

    The image below is of the Navy GOFS sea surface salinity field.  Enhancing this to focus the color bar on the salinities between 35 & 36 clearly shows the clockwise circulation of the warm eddy just southeast of the Mississippi delta.  Sally is right over this warm eddy now, still as a tropical storm in this image. NHC forecast says it moving slowly and they say that it still has time to intensify. Glider SG601 is on the northern side of this ahead, just ahead of Sally.  Kevin at USM is piloting it to the south to get into the eddy and even closer to the forecast track of Sally.

    Lets take a look at the temperature profiles from SG601 and compare to the models moves more into the eddy and continues to experience the winds from Sally.  Below the temperature profile is zoomed into the upper 200 m to get a look at the surface mixed layer in the upper 50 m where temperatures are in the 29C range.  Below 50 m, we move into the thermocline, cross the 26C isotherm about 65 m depth.  All models and data in good agreement down to about 75 m, then they diverge. GOFS (red) tends to stick with the Glider data (blue), while RTOFS (green) tends to be about 2C warmer between depths of 150 m to 200 m.


    Salinity behavior is the opposite.  Data and model are all in pretty good agreement below the mixed layer, but the salinity of the barrier layer varies widely.  Note that the temperature mixed layer is 50 m deep, but the low salinity barrier layer is only 15 m deep.  The salinity of the barrier layer is uncorrelated with temperature, so T-S relations don’t work, and you need to go out and measure the salinity. Light blue shows the spread in the salinity data from individual profiles.  GOFS in red looks to be in with the mix of salinity profiles, which is expected, since only 1 salinity profile is chosen for assimilation each day.  RTOFS has much less of a low salinity barrier layer.




  • “Rapid Intensification is a possibility…”

    Posted on September 13th, 2020 Scott Glenn No comments

    Here is a screen shot of the 5 am National Hurricane Center discussion.

    Last sentence in the screen shot tells us to watch the core of Sally on Monday. That rapid intensification is a possibility, ands stronger solutions like HWRF can’t be dismissed.

    Lesson learned from Laura.  Full text of the discussion here.


    Latest forecasts are below.  HMON rapidly intensifies nearly to Cat 4.  HMON is in the middle, intensifying to nearly to Cat 2.  Experimental HAFS stays a Tropical Storm.  50 knot spread in the intensity guidance.  Plenty of room for new science.