Joint Bayesian inversion of MT and CSEM data published

EM Lab graduate student Dan Blatter led a new paper on joint Bayesian inversion for marine magnetotelluric (MT) and controlled-source EM (CSEM) data to image submarine groundwater aquifers. The MT data is preferentially sensitive to the conductive parts of the model while the CSEM data is more sensitive of the more resistive low salinity aquifer. Joint inversion of the data yields the best resolution, as demonstrated with Bayesian uncertainty analysis.

Geophysical Journal International, Volume 218, Issue 3, September 2019, Pages 1822–1837, https://doi.org/10.1093/gji/ggz253

The paper is available for free here: https://academic.oup.com/gji/article/218/3/1822/5510458

Mapping offshore aquifers

Graduate student Chloe Gustafson and Professor Kerry Key published a new Scientific Reports paper on using marine EM to map large submarine groundwater aquifers on the US Atlantic Margin.  

Gustafson, C., Key, K. & Evans, R.L. Aquifer systems extending far offshore on the U.S. Atlantic margin. Scientific Reports 9, 8709 (2019) doi:10.1038/s41598-019-44611-7

The paper is freely available: https://www.nature.com/articles/s41598-019-44611-7

Selected stories from popular science news media coverage:

Morton, M. C. (2019), Huge aquifer imaged off the Atlantic coast, Eos.

EM Lab completes survey of Alaskan Subduction Zone

This has been a year for subduction zone electromagnetic surveys. In February we completed a large-scale survey of the Hikurangi Trench off the North Island of New Zealand, and now we just completed a complementary large-scale survey of the Alaskan subduction zone. Principal investigators Kerry Key and Samer Naif led  a large team that included Lamont graduate students Christine Chesley and Tanner Acquisto and postdoc Julen Alvarez-Aramberri  onboard the R/V Sikuliaq offshore of the Alaska Peninsula.  The electromagnetic and magnetotelluric imaging data collected using 159 seafloor stations and a deep-towed transmitter will be used to study fluids along the subduction zone plate boundary and how they impact plate tectonics and seismicity.

Click here to see photos and read the cruise blog.

Whales and sea otters galore

We’re on our way back to Seward with an ETA of 0800 tomorrow.  The ship is taking the scenic route via Shelikof Strait, which runs between Kodiak Island to the east and mainland Alaska to the west. Yesterday the crew had training drills planned so the ship pulled into the calm waters of Katmai Bay, and once again we got super lucky with the clouds parting as we neared the shore, giving us sunny skies and spectacular views of Katmai National Park.  We saw several whales, lots of sea otters and a few volcanoes.  After a month of working 24/7 on the ship for the EM survey, it was nice to have a day off while we steam back to port. In addition to enjoying the scenery, in the afternoon we had a round of 6 science talks in the ship’s lounge, including: analysis of melt inclusions from the western Aleutians (Janine Andrys), electromagnetic mapping of submarine groundwater off Hawaii (Eric Attias), EM exploration of the Middle America Trench (Samer Naif), EM mapping of seafloor gas hydrates  (Steve Constable), active seismic imaging of subduction zones (Tanner Acquisto) and aquifer systems extending far offshore on the US Atlantic margin  (me – Kerry).  I’m looking forward to another five science talks today, as its nice to get to learn more about what everyone is researching, including the students from other universities who’ve volunteered for our cruise, especially since we’ve spent a month together plugging in electrodes, cabling instruments, throwing grapnels, driving deep tow winches and going through lots of scotch  3M electrical tape, cable ties and all the things required to get new seafloor EM data at a subduction zone. Now we get to spend time learning about everyone’s research and ongoing projects.   Photos from Katmai bay below.

A little steam plume rising from Mount Martin volcano, as viewed from the RV Sikuliaq while in Katmai Bay

One of many sea otters we saw in Katmai Bay
Whale!
Eric stoked on the good weather and scenic views.

Steelhead is back

We pulled up to station 220 around 9 PM and sent a few release commands to receiver Steelhead but heard no replies. The Sikuliaq has a deployable centerboard that extends several feet below the hull and so we’ve been using the transducer on the end of the centerboard since that’s a less noisy environment for a transducer than up on the hull where more water turbulence and air bubbles can muddle the incoming pings. For almost all of the receiver recoveries that has been working really well.  But alas, we couldn’t get any commands through to Steelhead, though we did sporadically get a reply when trying to range on it so we knew it was still alive (in the acoustical communication sense). Earlier in the cruise Jake noticed another receiver that was hard to communicate with, but which worked just fine when we switched to the ship’s hull mounted transducer (as opposed to the centerboard ducer). So we switched to the hull transducer and bam – it acknowledged its release command on the first try! About four hours later we landed it on deck.  So new numbers for the cruise:  159 deployments and 159 successful recoveries. Coupled with the 168 deployments and recoveries we did on the HT-RESIST survey earlier this year, that’s makes for 327 total deployments without a single loss of instrument this year. To be fair, we did have two receiver frames and some sensors destroyed when the ship ran them over during their recovery this cruise, but we still got the data loggers (and the data) back.

Steelhead was back on deck around 1 AM with a nice turnout of the science party to celebrate the final instrument recovery.
Ocean bottom EM receivers stacked and ready to be craned off the ship when we arrive in Seward.

 

 

Last receiver recovered!

We did it! The last receiver has been recovered, giving us a total of 158 seafloor EM stations occupied in just about four weeks. Actually, we’re hoping this is the penultimate receiver recovery since we’re steaming back to station 220 on our first survey profile to see if we can persuade receiver Steelhead to release from its anchor. We had trouble communicating with Steelhead’s acoustic system and gave up on it back a few weeks ago, but now we’ve got some extra time on hand so we can drive around it to see if it can hear its release command better when the ship is at a certain azimuth; there’s a big seamount nearby that may be giving some reflections that muddle the acoustic signals heard by Steelhead and hopefully we will find just the right angle that allows it to hear its release command. Fingers crossed…

In the mean time, below are some photos from the last few days.

Our ship track since leaving Seward. You can see our three survey profiles that cross the deep trench of the Alaskan subduction zone. The shaded colors are high resolution multi-beam bathymetry obtained by lots of previous cruises in this region, but even so you can see there are still lots of gaps. When people say we know the shape of the surface of Mars better than we know the shape of the seafloor on Earth, these gaps are what they are referring to (and when zooming the map out, there are way more massive gaps in all the ocean basins worldwide).
A photo Jake shot of the RV Sikuliaq the other day when we were deep towing SUESI on the continental shelf. Jake went out with two crew members in the ship’s fast recovery boat to attach a few extra glass floats onto our towed EM receiver array located about 880 meters behind the ship and snapped this photo on the way back.
Jake taught the day shift how to make a locking Brummel eye splice in a line of amsteel rope and here he is watching Peter finish up tapering the end before its final tuck.
Day shift with the Alaska Peninsula in the background.
Day shift with the Shumagin Islands in the background.

 

Land Ho!

Yesterday was an amazing day at sea. In the morning we got to see a pod of Orca (YES ORCAS!!!!) swimming past the ship. Check out Steve Constable’s photo of the orcas here. We made our way to within about 10 miles of Unga Island in Shumagin islands group just as the clouds were clearing, giving us a scenic view of the Shumagin islands as well as the nearby Alaska Peninsula coast, including Pavlof volcano and Pavlof Sister. The clouds kept clearing throughout the afternoon as we deployed SUESI for surface electromagnetic transmitter towing across the shallow waters of the continental shelf (similar to surface tows we did in a paper published today about mapping offshore groundwater on the US Atlantic coast). As the evening set it, the skyline lit up as the sun set just to the right of Pavlof, blazing the sky into a spectrum of yellow to orange to purple colors.  Check out the photos below.

Panorama of the Sikuliaq’s back deck as we prepared to deploy SUESI, a surface towed receiver, and two GPS transponder floats.
Chris ready to start deploying the towed array.
A Vulcan three-axis EM receiver (front) being deployed along with two surface buoys containing GPS transponders (middle and back).

 

Pavlof and Pavlof Sister volcanoes viewed from 50 miles offshore.

Sun setting behind Pavlof and Pavlof Sister volcanoes.
The after glow of the sunset lasted for a few hours; even the night shift got to enjoy it at midnight – a benefit of being so far north and nearly at summer solstice (i.e. the longest day of the year).

Un mes en el mar de Alaska trabajando en el proyecto EMAGE

Julen Alvarez-Aramberri, a Lamont postdoc from Basque Country, an autonomous region of Spain, has been crafting a detailed blog about our cruise and life at sea. Check it out here: https://ciencilari.wordpress.com/

We’re nearly done redeploying all 39 receivers and will be towing SUESI on the continental shelf later this afternoon. At the moment there several small islands in view, which is nice sight after spending the last few weeks mostly out of sight of land. Some photos from yesterday below.

Jake attaching a mesh bag of styrofoam cups to a receiver being deployed in 4800 m water. Check back in a few days to see the shrunken cups.
Receiver being deployed in deep water.

Movie trailer

We finished recovering all 39 receivers and now are just about to redeploy them. We’re splitting the fleet in half, with twenty receivers to be deployed on the abyssal plain to collect MT data and 19 to be deployed on the continental shelf at 80 – 100 m water depths, where we plan to surface-tow SUESI so we can get shallow water CSEM data sensitive to the upper crust and  MT data sensitive to the crust and the deeper mantle wedge. This will extend the forearc slope data so that in total we will have a 160 km profile of CSEM and MT data from the trench up onto the shelf (plus about 50 km of CSEM and MT data from two crossing profiles on the forearc slope). And with the abyssal plain data, the MT profile will be about 240 km long. Although SUESI has a high pressure leak somewhere, we tested that it can still hold a vacuum and so we feel safe lowering it to 10 m water depth on the shelf (i.e., only 1 additional atmosphere of pressure).  The weather prediction is looking great for the next week so we’re all hoping we can successfully finish this last leg of the project.

In the mean time, here’s a funny EMAGE movie trailer made by Bern:

! – WARNING *** LEAK DETECTED ***

The last few days have been quite eventful. Let’s walk through the sequence with a series of photos and videos. You can also read more about the events on Steve Constable’s blog  at: https://marineemlab.ucsd.edu/Projects/Megathrust/index.html

First we deployed all 39 seafloor receivers along a forearc-crossing profile (aka up the continental slope) and got SUESI in the water and transmitting, all in less than a day. Since some rough weather and large swell was in the forecast, we decided to deploy all the receivers in the shallower depths of the forearc; that would pose less of a problem for deep towing since we’d need to let out less tow wire at those depths and hence would have lower wire tension spikes when the ship started heaving in the big swell. Then next week when the weather is better, we could move all the receivers out to the ~4.8 km depths of the abyssal plain, and hence we would be able to let out more deep tow wire in the gentler seas without worrying about exceeding the wire’s tension limits.  An upside of this is that we decided to deployed a mini 3D array of 24 receivers (4 x 6) deployed every 4 km, which would give us the first 3D CSEM survey of forearc structure.  Here is the night shift deploying SUESI after breakfast:

A clear deck means all the equipment is deployed and SUESI is transmitting current.
While collecting data, we saw SUESI’s output current drop from about 250 amps down to nearly zero, a telltale sign that we had either lost an electrode on the antenna or a leak had burned through one of the antenna’s cable terminations. Here’s Chris and Jake on the fan-tail with recovery hook poles as we bring SUESI back aboard.
Bummer – we did burn a cable termination. But luckily it happened right at the mounting point on SUESI, so we could relatively easily swap this with a spare antenna lead.
If the termination gets worked even a little bit it can increase its electrical resistance just a tad, but when you’re pumping 250 A through it, even a tiny increase is enough to cause it to heat up, which led to a runway process of the plastic insulation melting and electrical current leaking to seawater. We replaced the antenna lead and just as we were about to redeploy SUESI, Steve walks into the lab and says that Brandon spotted some loose screws on the A-frame joints! Thankfully the screws where just on protective plates covering maintenance access points, so there was no danger. One of the ship’s engineers came out and tightened them all up.
Away goes SUESI as we redeployed it  after looping the ship back up the tow profile a bit so that we wouldn’t have a data gap from where we had to pull it up due to the melted cable termination.
Here’s a look at the forearc receiver profile and grid of 3D stations. The bright cyan lines and ship icons show cargo vessels passing by along the great circle path that is the most efficient route from the US West Coast to Asian ports. Not exactly the best place to being deep-towing but they’ve been giving the Sikuliaq a wide berth.
As predicted, the rough weather arrived and the Sikuliaq started pitching and rolling a lot. Luckily the wire tension on the deep tow cable stayed low, so we were feeling good about deciding to survey at shallower depths of the forearc first.
It was getting quite boring, meaning everything was working really well, and then I saw this on the terminal that displays the telemetered data from SUESI’s onboard computer. ! – WARNING *** LEAK DETECTED ***   Yikes!   Hearts racing, we powered SUESI down since we didn’t want 250 A output current if there was a leak, even a small one. Then Steve said it *might* be a glitch, and so we should power it back up to see if the warning comes up again. So we did, and the warning went away. Several minutes later it came back. Steve called his engineer back at Scripps and we learned the command we could use to toggle the leak detector on and off. We came up with a plan: we would toggle the detector off and then back on every time that message came up. If the warning came back on immediately then we definitely had a leak; if it was intermittent then we either didn’t have a leak and the detector was glitching, or the leak was just a tiny amount of water.  Over the course of the next few hours the message came on only sporadically so we decided the problem wasn’t significant and we carried on with the deep tow. We had already completed towing from the shallow continental shelf down into the trench and then had completed the two deepest crossing profiles on the 3D array, so we were feeling good about that, but we were still feeling delicate about the transmitter. I went down to the lab to check in with Eric while he was watching SUESI’s backup terminal screen. That’s when I heard a loud pop noise in the aft lab where SUESI’s  top side power supply is located.  I went to check it out and noticed its voltage had jumped up from about 1680 to 1704 V, and then we noticed SUESI was no longer telemetering data back to the ship. Arrgggh!  So I made the proverbial “phone call to wake up Steve” and he came down to the lab. Pretty quickly he decided it was a problem with the power supply and not related to SUESI’s leak. Regardless, we still had to bring SUESI back on the ship since it would take time to fix the power supply. We were already using the backup power supply since we had blown a transformer in the first one earlier in the cruise, so we decided to beach SUESI and start recovering receivers.
SUESI being brought back up.
Here you can see where the 120 V to 2000 V transformer blew earlier in the cruise.
Old transformer out, new better transformer in. After opening up the other power supply, we  also found out that the popping noise I had heard was from a big fuse blowing on the part of the circuit that puts FSK telemetry on top of the 1700 V 400 Hz signal we put down the deep-tow cable.
The next day with SUESI back on deck and the power supply fixed, we powered it back up and sure enough, we were getting constant LEAK DETECTED messages, so we decided that this was no longer a glitch and was likely a bonafide leak. Dang it. Steve and Jake decided to try drying out SUESI’s insides. SUESI’s pressure case is too bulky and heavy to safely deal with on the ship so instead they opened up some of the seal screws and used a vacuum pump to move desiccated air through it while aiming a heat gun at the base of the pressure case to help speed up the rate of evaporation on the inside. After doing that for a long time, we fired up SUESI and it was no longer issuing leak warnings, so case closed – SUESI did indeed leak. Crap. That means it is no longer safe to deploy into the deep depths of the abyssal plain. And further, we were already using the backup SUESI since we had some significant corrosion issues on the first one after towing the first two profiles, so we are now thinking about contingency plans. The current idea is to deploy have the receivers on to the abyssal plain so that we can at least get passive (and much lower frequency) MT data there to look at lithosphere-asthenosphere boundary structure. Then we will take the other half of the receivers and deploy them up in the 80-100 m depths of the shallow continental shelf to collect MT data to look at the deeper crust and possibly the mantle wedge; we will also surface-tow SUESI’s antenna across the profile of shelf receivers so we can get shallower sensing CSEM data there. We’ve seen lots of water current noise on the receivers in shallow water, as well their compass data showing some of those instruments being spun by the water currents, so there’s no guarantee we will get useful data on the shallow shelf, but we still have a week of ship time left and as long as we have anchors available we will keep on collecting any data we can. Also to put things in perspective, despite the setback with SUESI’s leak, we have already collected a TON of super awesome CSEM and MT data. We’ve been able to get CSEM and MT data along two profiles spanning the abyssal plain and forearc and we towed the forearc on a 3rd profile in addition to collecting a mini 3D survey on the forearc slope. So with all the data we have in hand already, the EMAGE project is a huge success. It’s a bummer that we won’t be able to get more abyssal plain data, but we have enough data in hand that we will be kept busy analyzing it for the next few years.