The “Butterfish Smackdown”: “This could get a little nautical”

"2011-12-16 13:25:23 GMT"
The “Karen Elizabeth” headed east north east as Chris tried to set the net for a tow along the bank.  The gulls pumped their wings hard to continue sidesliping along our wake. Despite the wind they were still as full of their greed as their grace, crying out their expectation loudly for what we soon might just toss back into the sea.  But it was blowing too hard to set the net crosswind, so over the “Karen’s” loudspeaker, Chris told Denny, Josh and Mike to wait until he finished turning the boat up wind and sea.  I was in the bridge too, waiting to enter the time and location where the net would begin to fish, which along with other descriptions of the tow including its catch, I would soon send over our gliders tail through a satellite to the Ocean Observing System onshore.  The “Karens” bow plunged deep into the face of an oncoming wave.  The world disappeared behind the white water in the windows, then between the rivulets of draining seawater the white capping, broad shouldered waves and crystalline blue sky that always follow the passage of a front, began to reappear.  I’m sure my eyes widened just a little bit. But I am more certain I didn’t allow Chris to know that they did.  While we fished gannets soared updrafts from trough to wavetop on long white and black tipped wings.  The northwesterly was gusting to 40 knots and the seas were 10 to 12 feet.  We were still working and I had gotten my gale.

Mike Broniewski with a clean tow of allot of butterfish

Late that night, but really it was the next day, when the wind was much lighter and the sea was down, Chris picked his last station. I had gone down below to get a little sleep until midnight when we were scheduled to reach Chris’s waypoint.  When I woke up and climbed into the bridge it was 1 AM and he was still looking.  “Listen buddy, the habitat model needs a station before first light too” I said.  He kept steaming for minute, and then with noticeable agitation turned around.  After about 10 minutes more we set the net again.  I entered the time, longitude and latitude.  “Where are we?” I asked.  “In the middle of nowhere”, He replied disgusted he had to fish here. “South side of Alvin Canyon.  Right on the bank” He added.  After 20 minutes of towing along the 84 fathom isobath we hauled back.  I went down on deck to help weigh the catch that I assumed would be light.  When the net bag opened, a little over 3000 pounds of butterfish along light traces of a few other species poured out and filled the fishbox.  I sorted until I heard Chris slow the boat for his second tow.

On the left is the real time monitoring system showing the configuration and position of the net including real time depths and temperatures.  The monitors for the 50 and 200 khz fisheries hydroacoustics are on the right. The bottom is broken in the acoustics because we are setting the net to tow downslope.  

As I climbed up into the bridge I said  “I’m crying crocodile tears for you. That one was over 3000 pounds”. “Well I wanted 10,000 pounds” he said as he moved back to the winch controls to set the net again.  This time we were towing down the bank.  We started in 57 fathoms of water with a bottom temperature of 58 degrees. The temperature held steady until the water was 83 fathoms deep, then it began to fall to reach 55.8 F at 93 fathoms.  It was time to haul back.  The acoustics showed fish stacked up where the bottom depth was between 89 and 93 fathoms.  Like our earlier observation the fish seemed to be stacked up downslope on the cold side of a subtle bottom temperature front. His second tow also produced 3000 pounds of butterfish.  Chris told me that he usually doesn’t fish downslope or with his doors armed with temperature sensors, so the results of these tows surprised him.  He told me he had learned something on our cruise.  I am glad to know that because I learned a tremendous amount from him about the fish that overwinter at the edge of the continental shelf which he fishes from Cape Hattaras to the Canadian boarder as if it was a river running through his own back yard.  And I have just started scratching at the surface.

Heading Home

Over the last 24 hours, the crew of the Karen Elizabeth completed 4 stations. These 4 Stations fill out the outer and mid-shelf stations we need in the northern Mid-Atlantic Bight based on time of day. Since the model and fisherman have indicated the importance of time of day, it is important that we sample the same stations during the day and again at night.




Early this morning the F/V Karen Elizabeth completed Station # 26. Station 26 is the final sampling station of the trip. It was done in the middle of the shelf south of Martha's Vineyard. They are now steaming home and expecting to get back at the dock around 3:30pm this afternoon. We thank the entire crew for the incredible effort they made to make this a successful trip. They were able to get their work done on board and keep us all posted throughout the week.



As the Karen Elizabeth steams into Pt Judith, ru07 passes them to the west as it heads out to the shelf break. RU07 is an ocean glider like the one that has been strapped to the house of the Karen this whole week. Unlike RU10, RU07 is deployed as it is designed. It is an underwater robot sampling the the ocean from seafloor to surface as it heads along its route. Instead of sending files from ship to shore as was done with ru10, ru07 is sending back real-time ocean data every time it surfaces and makes a satellite phone connection to our lab. Here is the latest ocean data sent back from our ocean going robot.

Temperature:


Salinity:


Chlorophyll concentration (a measure of phytoplankton concentration)


You can follow along as ru07 makes its way south toward New Jersey at our website here: http://marine.rutgers.edu/cool/auvs/. Just click the name of the glider (ru07) to see the real-time data!

Unlike the remote sensing satellite and HF radar data that has been used so far in our model development, the gliders give us data on how the ocean varies throughout the water column. This data is a look into the future. It is these scales and structures that match closer to the tools that Chris and the rest of the crew of the Karen Elizabeth were using as additional indicators of where and when to fish. I look forward to learning from the fisherman and ocean scientists on how best to bring this data into the products!

So as one successful trip ends another begins. We continue to study the ocean in ways that have not been possible before.

Sharing our Story

Today was the second day of the MARACOOS annual meeting in Washington DC. MARACOOS is the regional component of the Integrated Ocean Observing System (IOOS) in the Mid-Atlantic Bight. The focus of the organization is to develop and sustain ocean observing and prediction based products in the Mid-Atlantic. The butterfish model that we are testing this week is made possible by the significant effort of this organization in providing both their input data and expertise. For more information on all the MARACOOS activities please visit their website here: www.maracoos.org.

Each year at this annual meeting, I learn about many of the exciting activities going on in our region. These activities are organized into themes: Maritime Safety, Ecosystem Decision Support, Water Quality, Coastal Inundation, and Offshore Wind Energy. In each of these themes we learn about the activities and needs for products and services.

In that context Greg DiDomenico, Garden State Seafood and the lead on our Butterfish project, and I had the opportunity to tell the story of this product development and evaluation as part of the Ecosystem Decision Support Panel. We thanked those that made and encouraged the partnership approach that we took, including NEFSC Cooperative Research Program and MARACOOS.

Our presentation went through the steps we all took to have the butterfish model ready for the evaluation this week. We stressed the critical need for the partnership between industry, academia and government in the development. Each group brought their own experience and expertise and the product we tested on the Karen Elizabeth this week was built on that partnership. Perhaps the most remarkable part for me was that the real-time link with the boat allowed us to include the discussion that John and Chris are having on the bridge, before they even get back to shore. We could share with everyone in the room the true sense of how this partnership is put into practice.

It is very exciting to work in this environment and all involved look forward to continuing this great partnership!

The Butterfish Smackdown: Fishing for Temperature

The three quarter moon glows dimly behind a thin unbroken layer of high stratus cloud. Tonight is nearly as black as any night can be.  I walked out onto the bridge deck, eye level to the net reels, to look out at the horizon. I thought I saw it, but it is just as likely that I made it up.  We are cruising up the bank to a nighttime station Chris has picked about 50 nautical miles east north east of the mouth of Chesapeake Bay.  We have already finished fishing model pixels representing “good” and “bad” habitat that we selected as randomly as possible given what we still need to get done before Sunday. The northerly wind has died down so the “Karen Elizabeth” has lost the heave she had earlier this evening.  The only waves to speak of are those she throws up which cap as ghosts in her wake.

A few hours ago we fished Chris’s daytime station with the net doors armed with real time and recording temperature and depth sensors.  He picked a bump along the wall of the shelf break he said is “sometimes pretty fishy.” The record of his trawl tows stored in his Navigation Software shows he’s worked this feature a lot.  His first tow followed the 114 fathom depth contour, and in real time the temperature at the doors held to a pretty steady 53 degrees.  The low frequency acoustic machines lit up with bright orange and red targets and the 200 kilohertz machine showed a lot of bottom haze.  “It might be hake‚” he said.  When we hauled back the net was indeed full of small hake, a few dogfish, and about 4 pounds of butterfish.

Chris then steamed up onto the edge of the bank into 55 fathoms of water, turned the boat east, and set the net for his second tow. This tow was down slope and in the shallow water the bottom temperature was 61 degrees.  The hydroacoustic screens were mostly black to the bottom. As we moved down the bank the temperature held pretty steady; at 80 fathoms it was still 59 degrees.  Then all of a sudden at 90 fathoms orange and red targets began to slide into the acoustic screens, with the 38 kHz and 50 kHz machines showing hard red targets.  The animals kept showing on the screen as we towed a few more minutes longer.  Then the temperature began to drop 58, 56, 55 then 53 degrees as the bottom fell away under the boat to 140 fathoms.  A huge mass of fish appeared on the acoustic screens that our net, with its sensors a quarter of a mile or more behind us, never caught.  Our 20 minute tow was up - we had to haul back.  The numbers of hake, dogfish and other species caught in the net didn’t tell the story.  We plotted up the data from the temperature depth recorder and could see the abrupt drop in temperature as the netdoor reached 90 fathoms.  It showed we had pulled our net just barely past the wall made by the warm water on the bank and the cold water down on the other side where Chris’s first tow showed fish were abundant.  I put the position of Chris’s station on the map of our habitat model “nowcast” we made made with measurements of the sea surface from satellites in space, radars onshore, and the broad scale seasonal survey of animals in the Northwest Atlantic NOAA performs twice every year.  Chris had sampled a hot pixel in our habitat map.  But he had also drilled down to scratch the surface of much finer scale processes and features, like that deep water thermal front, that are embedded in broader scale features and affect the lives of fish and the fisherman who depend on them.

Turning North

The Karen Elizabeth has turned back north after completing two daytime stations in the southern Mid-Atlantic Bight. The next stations will fill in gaps in the two regions already sampled off Delaware Bay and Martha's Vineyard based on time of day.


Here is a Station 16 report from the ship:
Chris's station late this afternoon. This was in the hotzone but he didn't find that out till afterward. The acoustics screen shows the type of stuff we saw on the screen at depths greater than 165 meters.



The temperature and depth plots below are from the doors. The second tow was perpendicular to the bank. At about 160 meters, we hit a temperature front and the fish started to be visible in the cold water pressed up against that warm water boundary.

This is the kind of fine scale stuff going on inside our great big butterfish pixels that are affecting the fishes.

The Butterfish Smackdown: Great People and Good Probes

Another blog post John sent from the ship:

Josh boarding up the wall of the fishbox.  The wire lifting the net is the whip.

We have maxed out our sensors to better measure the fluid in which the animals live...or don’t live.  What we have right now is pretty simple, consisting of recording and real time temperature/depth sensors. Mike Ball, my main man from NOAA cooperative research, has brought several recording temperature/depth sensors.  As an aside, Mike has been tirelessly working up the catch with the “Karen Elizabeth,” supercrew, confirming species identifications, measuring animals and recording data. I would be helpless out here without him. He has made me look very good.

Mike and Denny at the fish conveyor belt.

I have asked him to reset the sampling rates on the temperature/depth sensors so they make a measurement every second.  We now have one sensor on a net door and download the data after completing each station.  Chris has asked that the other probe be put on the “headrope” at the top of his net.  The headrope is equipped with floats that help keep the net open.  That sensor combined with the bottom depth will tell Chris how high the net is and thus the thickness of the bottom layer it is fishing.  This sensor will also give us the temperature at a slightly higher level in the water column.  Chris has also put a sensor on one of the net doors that reports bottom temperatures in real time to us in the wheel house of the “Karen.”   With the data from these sensors we can begin to analyze the temperature and depth preferences of the animals we are collecting pretty precisely.  I am now sleepless, thinking about the battery of other sensors I’d like to fix to Chris’s net doors in order to measure the properties of the fluid on the next trip out if there is one.

Temperature-Depth measurements taken from the doors of the net for tows a few days ago.  The first is for a tow where the model predicted bad butterfish habitat, the second is for a tow location chosen by Chris, and the third is for a tow where the model predicted good butterfish habitat.

Our sensor data is useful for another reason.  There are many things we need to do to refine our iterative approach for developing habitat models for continental shelf species. One improvement is to couple our habitat models to physical oceanographic models.  Yes this will allow us to evaluate and explore dynamic habitat models in real time when there are clouds between the ocean and the satellites.  But more importantly, moving toward oceanographic models will allow us to consider the properties and processes of the fluid below the surface of the ocean where the fish actually live.  This kind of approach has been used in the Pacific Ocean to model the habitat of tropical tunas.  But we need to evaluate the accuracy and precision of oceanographic models, just as we do our habitat models.  We have begun to send our temperature records over the satellite telephone in our robot glider's tail back to Dr. Avijit Gangopadhyay, an ocean modeler at UMASS Dartmouth.  Using our data he can begin to compare our real time measurements below the ocean's surface to the subsurface temperatures in his simulated ocean.  And he can begin to help us think about the ways we can use his models to inform ours.

South toward Norfolk Canyon

After a successful repair in Cape May yesterday, the Karen Elizabeth headed south east toward our southernmost hotspot in the buterfish model. Along the way they stopped at a mid-shelf station (Station 12) off the coast of the Delmarva Peninsula (below).


Here are some of John's notes on Station 12: 'This is a cold habitat station midshelf. We now have 3 temperature depth sensors on the net. One on each door. One has a live feed to the surface, one is recording with a sampling rate of 1/sec. Chris has added a second recording temperature/depth sensor to the headrope to measure its height over the bottom.'

After finishing Station 12 last night, they continued south to the hotspot off the mouth of the Chesapeake Bay. Stations 13 and 14 were done at about 3:00am and 6:00am local time. In all three stations conducted last night, the tows came up empty of butterfish.

In addition to the ocean environmental variables that go into the butterfish model, there is a time of day term. This is quantified as the sun's elevation relative to the horizon. We see a strong day/night variation in the model that has been confirmed by the fisherman. Below is the same area of our model with a prediction of butterfish habitat during the day. The previous image above is the same region at night.


You can see the hotspot surrounding the Karen Elizabeth light up during the day in the model. It will be very interesting to see how the sampling and acoustics today differ from last night's.

Clear Skies, Dec 13th

Finally, some good coverage and clear skies over the Mid-Atlantic from MODIS-Aqua! This may be our last shot for a while, so I hope it was a good one. The data just posted to our site, so we will see soon what kind of butterfish prediction we get.

The Butterfish Smackdown: “Coming Home to Roost”

Here is another blog post we got from John this afternoon...

Well not home actually but to Cape May, New Jersey for a few hours.  We are steaming in, at just above an idle, from Spencer Canyon where we fished yesterday afternoon.  Chris wants to reach the Cape May harbor entrance at first light. One engine is running too hot and we are meeting a mechanic at the dock.  Chris hopes the problem is with an engine fuel injector, a quick fix and turnaround.

Yesterday afternoon we started sampling our new plan on a cold spot on the  butterfish habitat model about 13 nautical miles west of the shelf break near the canyon.  The clouds in the satellite data didn’t give us a lot of pixels to work with, but they did give us enough. We also have the previous days’ predictions and the 2010 model hindcast movie.  All three show this area light up, the movie shows the dynamics, and the fishermen confirmed the pattern before we left the dock.  Our cold spot produced no butterfish.  I gave Chris the option of sampling his station before or after the station in the hot pixels from the model.   He decided to go next. 

We headed due east about 9 nautical miles toward the shelf break, but turned to run south along the 50 fathom line.  Chris looked back and forth from his hydroacoustics to the water, which was calm in the light breeze. The water began to get streaky, the kind of thing you see when you are sitting on the convergence between two water masses defined by temperature or salt and the current is flowing in slightly different directions causing the water to shear.  It’s the kind of thing Josh, as a physical oceanographer, understands much better than I do.  I wish he were onboard here too. 

Patches of telltale butterfish “dust” began to appear just above the bottom on the monitor of Chris’s 200 kHz acoustic machine.  “I don’t know what this will produce, but we’ll try here.”  It was show time for him in this smackdown and he looked a little nervous to me.  We spun around and set the net.  As we towed along the 50 fathom isobath we talked about his observations of the movements of fish and squid in relation to the continental shelf break, slope, and canyon walls and the movements of water masses along them over the course of the winter.  Our 20 minutes was up. It was time to haul back.  The bag was so heavy, Josh Wallace, one of three “supercrew” deckhands on the “Karen Elizabeth‚” had to wrap the “whip‚” the wire off the center crane, around the net bag before it reached deck level so it could be lifted aboard. Then Denny Stamand, supercrew number two, tripped the net into the 8 x 8 x 3 foot deck “box” which filled to the brim with a very clean catch of butterfish. They shined like silver dollars‚ 2035 pounds of silver dollars to be exact.  As Denny and Josh got the totes and baskets together to begin to sort the catch from the conveyor belt table, Mike Broniewski, supercrew three, secured the steel doors that fly like wings underwater and keep the net spread. “It’s a few,” Chris shrugged and then swung the boat around toward the south and the start of his next tow that is set by our sampling protocol to be 90 degrees to the first tow and thus “downslope.” 

As we steamed to the site he talked about really big “catastrophic” tows of butterfish down off Oregon Inlet during the fall - “how the fish settle down so tight and thick to the bottom in the late afternoon you can’t see them on the 200 kHz machines.” From his description it sounds exactly like those shape shifting flocks of tens of thousands of birds that curl like smoke in the autumn sky until one bird decides to roost and the rest follow all at once. In 10 minutes 20,000 lbs of butterfish can fill the nets so full the fishermen couldn’t get them on deck even if they wanted too.  It can take well over an hour to dump the nets and recover.  That can kill the squid fishing for the day.  Chris’s second haul produced an equally clean catch that included 2386 pounds of “silver dollars.” 

Now it was the model’s turn to define habitat, by using satellites on land and radars on the shore, bottom depth and a few other things.  We picked a random pixel in the hotzone 11 nautical miles to south and slightly to the east in 55 fathoms of water.  It was close enough to the shelf break that that our second cross isobaths haul took us down the bank.  Each haul produced 4 lbs.  This was pretty embarrassing and no contest in that “smackdown.”  But should it be embarrassing?  Both Chris’s and the model tows were pretty close together in space.  Our model is pointing us in the right direction, giving us a broad scale view of the dynamics of the environment from the fish perspective, given surface ocean features we can measure using radars on land and satellites in space, and a broad scale seasonal snapshot of butterfish abundance on the continental shelf measured from Hatteras to the Gulf of Maine by a single ship.  But within these broad scale dynamic features, there are finer scale dynamic gradients, the fish and fishermen are using as cues that are too fine for us characterize from space and a single coarse scale characterization of the continental shelf ecosystem.  Habitat is dynamic, defined by suites of particular environmental features and processes that vary at different scales of space and time and thus nest in time and space to define ocean habitats.

I felt pretty down after losing last night’s smackdown, until I realized the power of our new sampling approach.  Imagine if we had 2 or 5 other fishing boats working with us equipped with high tech sensors, focused on the area around and including each hot area defined by the model we made with the fisherman.  Let the best fisherman we can find define at least 1/3rd of the stations in each area.  We can then sample the balance of stations in a stratified random manner in both good and bad habitats, so that we can make contrasts and identify conditions fisherman and scientists might have both missed.  How exciting would that be - a high tech pirate navy for process based habitat research.

Looking for a clear day

Our latest cloud top map measure by GOES-13 and received by UD's geostationary satellite receiving station indicates that we may have good weather today for the MODIS-Aqua measurement we need at 1:00pm local. I'm hoping that we get a great shot so that John Manderson can get some great Butterfish model predictions that can be tested by the fishermen. All this technology, and we are still subject to the weather.

Tough day for Satellites

Well, the clouds moved in with force today, covering our study area just as NASA's MODIS-Aqua flew overhead of the University of Delaware satellite receiving station. Matt Shatley at UD has configured our system to us the Naval Research Lab Automated Processing System. The ORB labs machines convert the output to a netcdf layer and push the MODIS-Aqua data publicly on a THREDDS server for anybody to use. The MODIS-Aqua data lives on the Amazon cloud. Kyle Wilcox of ASA configured the cloud server to handel these large data files and make them freely available. Then, we crunch this data and map it on Google Earth to bring it all together in one place. This is just one tiny thread of the technological tapestry being woven by integrating universities, businesses and governmental entities. Looking back at the blog entries for this experiment, I am amazed at how many organizations make this possible: NOAA, NASA, NMFS, Navy, ASA, Amazon, Google, Garden State Sea Food, local fishermen and universities. I'm sure there are some I have forgotten. If you like working across all of these entities, check out MARACOOS.

Butterfish Smackdown: Stray dogs

Here is a blog post that John sent us from the ship this afternoon.


Chris moved from the wheel to the winch controls that face aft toward the net reel to set up the first tow at the last station we will sample to the north side of the Hudson Shelf Valley till we return from the south. As he crossed the wheelhouse he glanced at the fishery hydroacoustic screens and the arcs of green, yellow and red light against the black backround on his 38 khz machine. 'Looks like a pack of stray dogs', he said. Sure enough we hauled up a thousand pounds of spiny dogfish less than a foot long after only a 20 minute tow.


All day Sunday I tried to pull Chris back from the offshore edge of the footprint of our model. All day long he kept pushing harder to get us over the edge of the shelf and onto the dropoff. This was a real problem because we made our butterfish model using the best available data. This included not just the satellite and HF radar information but also the NOAA fisheries survey data. And the fisherman we talked too told us that the movie made of the model we built with their help looked pretty accurate.

The NOAA survey is remarkable. It has been performed twice a year since 1963 and covers the continental shelf from from Hattaras to Canada. But doesn't cover areas in the nearshore including estuaries, or habitats off the edge of the continental shelf. So our butterfish habitat model is really aimed at habitat conditions on continental shelf, not off the edge. I wrote a note to Laura, Josh and Matt apologizing for pushing the outer boundary so hard and promised to pull us back to do some midshelf stations. To truly evaluate this model we need to stay on the shelf.

But wait, what if Chris was telling me something about butterfish and their habitats I didn't know? What if really prime butterfish habitat during this time of year isn't on the shelf at all but just off the shelf break. Since this is a habitat modeling study we really need to understand the full range of environmental conditions where the animals live. And we already have the NOAA survey performed this fall to evaluate our model for the area of the shelf area. Why not change up our research plan and adapt quickly to what we have learned over the first 36 hours of this cruise. Like 'stray dogs' lets systematically spend a little time off the reservation and evaluate our model footprint too.


So here's the new plan. We are still going to sample areas where our model predicts there is potentially good and poor butterfish during the day and the night off the Chesapeake and Delaware estuaries. But in each of these survey areas we will give Chris a station during both times of day to use his practical ecological knowledge of butterfish to show us what we don't yet know.

The Butterfish Smackdown: The contributions of Mike Ball from NOAA Cooperative Research

Mike Ball, my deck hand, at the sorting table

Our field work would not be possible without the contributions of Mike Ball.  Mike works for NOAA Cooperative Research which has generously provided the support for our project.  Mike has spent a lot of his time working with commercial fisherman and on commercial boats in New England and is an expert at identifying deepwater fish.  I, on the other hand, have never been on a deepwater boat and don't know my deep water fish all that well.  Mike has led all the deck work on this project and as a result it has been going very smoothly.  We are very grateful for his effort and enthusiasm.

One more Station and a Revised Plan

Station 8 was completed an hour after midnight local time. They sampled the outer shelf just north of Hudson Canyon. They are now heading south toward the hotspot in the model southeast of Delaware Bay.


The plan moving forward will be to complete 6 stations in the area indicated with a red marker above. Two stations will be inshore of the shelf break in a region of poor butterfish habitat predicted by the model, one in the day and one at night. The second two stations will be in the region of preferred butterfish habitat predicted by the mode out closer to the shelf break, day and night. The final two stations will be chosen by Chris Roebuck, our captain, as a region of good butterfish habitat predicted by the model constantly running inside his head. These final two stations will also be done once during the day and once at night.

Thanks to the reliable connection we have with the ship, we are getting some great information back and forth between ship and shore. Here are a couple samples of the kind of feedback we are getting from the ship.

Here is an image that John sent us last night showing a pile of fish just after a station. The image is a little blurry to keep the file size down, but we can get a feel for the work being done on board.


Several times a day John is also sending us messages. He is giving great insight into how they are sampling and adapting to the local conditions, real-time information coming from the ship's acoustics, and the model. This is a quote from him describing the great night they had last night:

'2011-12-12 04:21:25 GMT: Beautiful sunset and fickle winds as we made our last tow at station 7 just inside the fishtail. Chris and I were talking about the acoustic signatures of fish as below the crew sorted over a thousand pounds of spiny dogfish less than a foot long. Arcs of yellow occurred throughout much of the water column in the screen of the 38 khz hydroacoustics as we set out the net and he said "Bet that's a bunch of stray dogs." As he moved away from the console of controls for the crane and winches he said "the longer the stretches of good weather like this the worse the gale you can expect." He turned, looked me straight in the eye and cackled as he moved back to the wheel to steer us toward the Hudson Canyon.'

Daytime Samping Along the outer shelf

Sunday was a busy day for the Karen Elizabeth. Since this morning 4 additional stations were completed. Three during daylight hours and a forth finished just an hour ago under darkness. All four stations are out near the shelf break. In the map below the stations are indicated by a blue (nightime) or yellow (daytime) marker. Again the latest ship location is shown as the glider tail labeled 'ru10'.


Of the seven stations completed so far on this trip, the only one with more than 110 lbs of butterfish landed was Station 5. This station is located well offshore in an area just outside of our model coverage. The gaps in the butterfish model in this area are due to the presence of clouds in the satellite imagery that go into the model. Matt posted some good news today with a clear sky over much of the Mid Atlantic Bight.

Here is the latest prediction of butterfish habitat.



Next the Karen Elizabeth is going to heavily sample the shelf on its way south to our next area of focus off the Mouth of Delaware Bay. From the map, you can see an area of preferred butterfish habitat at the shelf break with a large area of blue (poor butterfish habitat) inshore.

Clear Satellite Image for Dec 11, 2011

Finally some nice, cloud-free coverage of the Mid-Atlantic Bight from NASA's MODIS-Aqua platform. The ocean sure looks a lot calmer from space than it does from a deck of a fishing boat! There are still some white puffy clouds offshore, but hopefully those will clear in the coming days. Because cloud cover can be a problem in the winter, our strategy for using satellite data has been to use 8-day rolling averages. Today's satellite pass should really enhance these composites in our study area. Once our machines are done crunching the satellite data for our Butterfish model, we will have a look to see where we find the latest ecosystem gradients and update the Butterfish model nowcast. Our attempt to merge bits and bytes are merged with grit and nets is underway.

The Net Hits the Water Under a Butterfish Moon

Last night the crew of the Karen Elizabeth was busy sampling the mid-shelf south of Martha's Vineyard. Soon after they left the dock they did a short station we call a 'shakedown' station. This allows the ship and science crew to go through the logistics of the work needed to complete a station before they get to the critical stations inside the footprint of the model. It is a great chance for everyone to see how a station will work and what adjustments should be made before the next one.

For this trip, a station is two short 20 minute trawls. One trawl will be running north/south and the second will bisect that line running east west. Once the nets are back on board, the crew will measure the total weight of butterfish (in pounds) and record the distribution of butterfish size for a random sample of 100 individual fish. Here is a picture of the large nets that are towed by the F/V Karen Elizabeth.


Here are the ship's winches that haul in the net after the trawl.


The glider strapped to the boat is calling our lab once and hour. With each call we get an updated position and an opportunity to share data from ship to shore. We are sending daily updates of the model every night to the boat and John Manderson is sending back the catch results of their sampling stations as they are completed.

Here is the position and track of the Karen Elizabeth as of 5:00am local time this morning. In all the maps we show during the trip, the location of the Karen Elizabeth will be shown as a glider's location, labeled 'ru10'.


You can see the path of the boat moving south out of Pt. Judith, RI, around the eastern side of Block Island and offshore. The three most recent red dots at the end of the track indicate where the stations were done last night. These stations were at 9:00pm, 2:00am and 5:00am local time. It was a very busy night on board.

The background color is the prediction of butterfish habitat updated for 1:00am local time. Once again the blue areas indicate habitat predicted less suitable for butterfish and the red areas are preferred habitats for butterfish. As you can see from the map, the stations last night focused on the mid-shelf in regions of relatively poor butterfish habitat. According to the model, each station is moving closer to more favorable habitat.



Today will be our first look at the mid-shelf during the day. Above is an afternoon shot from the model. You can see that the yellow areas of the 1:00am image are now orange. Since our model is an indication of when the bottom habitat is preferable to butterfish, this indicates that according to our model, the bottom is more habitable to butterfish in the afternoon than at night. The Karen Elizabeth is now heading right to the hotspot in time for a peak mid-day sample!

Game plan for 'Butterfish Moon'

I read the post that John wrote earlier today and I really like the mission name we have adopted for this trip, "Butterfish Moon". The F/V Karen Elizabeth (below) left dock today for an 8 day trip in the Mid-Atlantic Bight. They will be guided by their knowledge of the ocean, the capability of the boat, and of course the butterfish habitat model! I am very excited to see how much we will learn about the ocean and its ecosystem in the next 10 days.




The latest butterfish model for December 9th shows several regions of preferable habitat for butterfish. A large region South of Martha's Vineyard stretches across the entire shelf. Further south there are two areas along the shelf break off the mouth of Delaware Bay and Chesapeake Bay (below).


We want to make sure that we sample the high and low areas of the model. We have set the game plan to focus on both the mid-shelf and outer shelf. The first full day of the trip will focus on the 'hotspot' south of Martha's Vineyard. Under the butterfish moon they will sample the midshelf in a region where the model is predicting more suitable habitat for Butterfish. They will spend about a day and a half in this region going between the mid-shelf and shelf break.


We spent much of the last two days looking closely at the environmental data that went into the model. Avijit Gangopadhyay is an ocean forecaster at the University of Massachusetts, Dartmouth. He has been tracking a large warm meander of the Gulfstream that is right up against the shelf, south of Martha's Vineyard. The great image from Jennifer Clark, above, shows the meander of warm water from the Gulfstream getting very close to the shelf break. Jennifer is an expert on Gulfstream oceanography. Check out her site here: http://users.erols.com/gulfstrm/. Thank you Avijit and Jennifer for contributing to the blog.

This meander is likely impacting the habitat characteristics over the immediate shelf. We think that part of this impact is seen in the preferred habitat identified in the Butterfish model. The Karen Elizabeth will be there tonight taking a much closer look!

The Butterfish Smackdown: "Butterfish Moon"

The "FV Karen Elizabeth" with the glider strapped to her head

“We usually get two gales in a 9 day fishing trip in December” Chris Roebuck assured me with a slight trace of a smile.  Josh and I were taking a short break from networking our two laptop computers and the satellite telephone in the tail of the robot glider we had just finished lashing to the bridge of the “Karen Elizabeth”.  Meanwhile the crew was busy offloading 40,000 lbs of boxed and frozen squid onto pallets they then moved with a forklift to the shoreside freezers.

The crew offloading frozen and boxed squid as it streams out of the hold of the "Karen Elizabeth"

Two electronics technicians were also busy in the bridge integrating several new sensors for fishing and navigation into the ships computer system that surpasses anything I have ever seen on land, let alone the water.  We felt pretty ridiculous standing in the same space with them struggling to get our two little laptops talking for the third time.  Seems we have lost the “smackdown” in computing technology too.  But I was able to get access to the gliders memory and satphone.  We are not sure how we did it, but we did. And now it really matters. 

The computing "Smackdown":  Our two measly little laptops at the top, and below the battery of computers and computing screens that Chris has installed in the bridge to give him information on the position of the boat, the fishing gear, the fish, and future conditions of the ocean and atmosphere 

We downloaded the latest habitat prediction and revised our sampling plan with Chris’s input.  He’s sure we can fish more stations than we planned to and that’s good.  Tomorrow at noon we to set sail for the location due south of Martha’s Vineyard that our model indicates is a  hotzone.  Tonight on the Full moon in December we will make our first tow.  Chris told me last night the we'll be fishing our first day on the what the fisherman used to call a "Butterfish Moon".  That makes us feel lucky. 

Nowcast of butterfish habitat from the model we made with the fisherman.  Point Judith, indicated by the blue dot, is the port where the "Karen Elizabeth" is tied up till noon.  

Matt Oliver sent us a beautiful true color satellite image that shows the band of high pressure between the clouds that are gradually moving east.  This should give us a number of really nice days for fishing.   However that gale Chris mentioned might also be looming in the southwest 4 to 5 days away.

MODIS satellite image showing band of clear high pressure sliding in toward us to give s few days of really nice clear weather.

Actually, Chris and his wife very generously treated me to dinner.  We had a great discussion about the ecologies of butterfish, squid and other species, the fishing business, and continued to develop an interesting approach to the sampling that will begin tomorrow.   As I said my goodbyes we talked about the weather.  “Oh we’ll get that gale,” He reiterated with a grin.  Actually if it’s a well-behaved one I’m secretly looking forward to it.  But that’s awfully easy to say while still tied up to the dock.


Josh, Matt and Laura are now going to take over blogging while we are offshore.

Josh lashing the glider to the "Karen Elizabeths" head.

The "Butterfish Smackdown" VII "Uh Oh! This afternoon I leave for Rhode Island........"

What I saw in the distance from my bedroom window this morning.

A gull hung nearly motionless against the blue; teetering only slightly on its wings to keep its balance on the winds high wire.  But wait, I was still at home.  I blinked twice to clear sleeps last confusion from my eyes. Outside the wind was gusting to 40 knots and the air temperature was 40 degrees at the Ambrose buoy off the entrance to New York Harbor.  Outside my bedroom window and behind the Sandy Hook Peninsula, the sea was grey except where it rolled over white which was nearly everywhere.  Uh Oh! This afternoon I leave for Rhode Island, the “Karen Elizabeth”, and our “cruise” to the edge of the continental shelf to test our dynamic habitat model.


I hurried to pack the last few things; some books, a hat, a scarf and gloves, and my survival suit and headed off to Rutgers for the last bit of bench testing and our final preparations.  Laura, Josh and I met first thing to develop a “playbook” for adaptively sampling the dynamic habitat model.  Matt Oliver, our satellite oceanographer, skyped in from the University of Delaware to help, and we reviewed the last few days habitat predictions along with the satellite and HF radar surface current measurements that informed them.


Conditions are unseasonably warm in some areas of the mid Atlantic Bight for this time of year, and our model runs showed a hotspot still south of Martha’s vineyard, as well as east of Delaware and Cheseapeake Bays in locations similar to those we observed in Fall 2010 hindcast posted below.  So we have decided to fish standardized tows in habitats predicted to be “good” and “bad” in those 3 areas.  However, because the suns elevation (ie time of day) seems to make such a big difference, we have also decided to fish the “good” and “bad” habitats in the three areas during both the day and night so that our analysis is balanced.  We are pretty sure this will take less than the 7 days we have allotted to us on the “Karen Elizabeth”.  So on the way back up north, when we are not sampling model evaluation sites, we hope Chris will fish as efficiently and as effectively as he usually does.  We think we can learn a lot about the ecology of butterfish, squid and other animals by listening to him and watching him fish.

Josh Kohut, myself and Laura Palamara infront of todays butterfish habitat nowcast overlaid on sea surface temperature in google earth on flatscreen wall monitors in the Rutgers Coastal Ocean Observation Laboratory, RUCOOL.  Also known as:  "Mission Control for the Ocean"

After putting together the outline of our adaptive sampling “playbook” we did one more bench test of our communications and prediction system.  This included putting the most recent butterfish habitat nowcast up on the 8 foot by 10 foot flatscreen monitors in the Coastal Ocean Observations Laboratory.  We stood in “mission control for the ocean” for at least 20 minutes lost in a discussion about what was driving the dynamics of butterfish habitat based on google earth overlays of the daily satellite images, HF radar surface currents, and our habitat predictions.  We could have spent hours overlaying data on that big screen discussing the way the ocean and its animals work. We probably will when this field experiment is over.  But it was getting late and Laura helped Josh and I pack up the equipment and get in the van bound for Rhode Island and the “Karen Elizabeth”.

Daily sea surface temperature anomaly for the North Atlantic showing relatively warm temperatures in coastal waters from Chesapeake bay north into Canada.  These data are available here.

The “Butterfish Smackdown” VI. Geeking it up over satellite telephones & solutions for the transfer of data

Bench testing our communications system yesterday for our field experiment that starts this Saturday.  Laura Palamara makes a butterfish habitat prediction as Josh Kohut looks on wondering where in the ocean all the butterfish will be.  Laura's prediction map is sent via satellite phone to the glider David Aragon is wheeling out the door and onto the lawn at Rutgers IMCS.  The data is then sent from the gliders satellite phone to the blue computer on the right which is partially out of view.  This computer can be accessed by the other computer on the bottom right.  The glider, the two computers on the right and I are going on the squid boat,  the "Karen Elizabeth".  Josh, Laura and David will run the operation back at the lab along with Matt Oliver from University of Delaware and Steven Gray at University of Hawaii.

The Slocum underwater robot glider whose communications system we plan to use to transfer data from ship to shore and back again in our field evaluation of the butterfish habitat model we made with the fisherman.


There are a dozen reasons not to do this experiment. Actually we have already thought of a dozen that might cause us to fail and we haven’t even left the dock.  But we just keep moving forward because at this point failure is not an option.  Putting a regional scale habitat model into an operational nowcast mode is crazy in the first place, even if it is just experimental.  But this adaptive test of our statistical hypothesis, which is what our model really is, presents a formidable communications challenge.  It requires sending the model predictions Laura will compute onshore to the squid boat offshore near the edge of the continental shelf.  We need to navigate the squid boat on top of these model predictions so we can accurately position our trawl tows in regions predicted to be “good” and “not so good” butterfish habitat.  These predictions will change daily just like they do in the  "hindcast" movie in the last post. Then, after each tow of the trawl net, we want to send the results back to shore in “real time”.  Like everybody else, we take staying “connected” with cellphones and the internet for granted.  So the technical challenge of getting our data 100 nautical miles offshore and back hadn’t really dawned on us until just 10 days ago. But “necessity is the mother of invention” and we have come up with a characteristically unconventional solution.


We are going use a MARACOOS underwater robot for tracking the vessel and as a satellite data transmitter and receiver during our experiment .  Yes that’s right.  We are going to strap a robot glider to the bridge of the squid boat and pass the data back and forth over the iridium satellite telephone nestled in its robotic tail.  The glider is programed to send its position back to shore every hour so the lab can use this feature to track our squid boat in real time.  MARACOOS already routinely sends commands and data back and forth between the laboratory and robot gliders at sea.  So if we can set up that same kind of communication between the glider and a computer on the fishing boat our problem is solved—the glider becomes our R2D2 “operator” for our satellite telephone system.  Thanks to David Aragon and Chip Haldeman of RUCOOL we now have a laptop computer with a freewave wireless antennae and communications software that can “talk” to the glider when lashed to the boat. I have another laptop wirelessly networked with this one that we will also bring with us to fish offshore.  We will use this machine to write and analyze files including the model predictions Laura sends from shore through the glider.  Chris Roebuck, the captain of the boat, and I can also use this shipboard wireless network to send our catch data back to shore.  Based on the suggestions of Hamish Bowman, the marine GIS guru who lives in New Zealand, and who we met at last years MARACOOS annual meeting in Hoboken, New Jersey of all places (now this truly is a fortuitous and loose network!), Chris Roebuck and I will be use the Open Source Quantum GIS software to navigate the "Karen Elizibeth" on the habitat model predictions and choose locations to fish.  Like every other part of this project the solution to our data transfer and visualization of model predictions depended on the generosity and advice of many other people.  This is truly a collaborative, open source, research project.