MBARI Ridges 2005 Expedition
Juan de Fuca Leg: August 7–18, 2005
Gorda Leg: August 22–September 2, 2005
August 28 update posted by Dave Clague
Tiburon dive 889, Sediment sampling in Escanaba Trough - South transect
Today we made our third dive at Escanaba Trough, collecting another 30 pushcores to determine the distribution of glass particles from an eruption on the east side of Central Hill that produced a roughly 10-square-kilometer lava flow. This is the only young lava flow in Escanaba Trough, so the glass particles produced when it erupted are the only ones in the upper foot of the sediment. We have now completed dives on the north, west, and south sides of the inferred eruption site, with a fourth dive to the east planned for tomorrow. Our fifth and final dive at this site will investigate the inferred eruption site on Central Hill and collect sediment and rock samples near the vent.
Today we began our dive on the main lava flow and worked southward, traversing back and forth across the contact between sediment and lava. The edges of the flow were not located exactly where we thought they would be, as explained in the contribution below from Stephanie on sea-bttom navigation. With our new observations, which located the actual contact, we will be able to make some adjustments to the previously collected sidescan imagery, so that it more accurately shows the location of the flow.
We crossed several lobes of the lava flow during our dive today. These sheet and lobate flows appear to have erupted rapidly, forming shallow lava ponds that subsequently drained as the flows advanced beyond them. In several places, we observed evidence of such pond drainage, which leaves behind bath-tub rings of lava on the edges of the former pond.
We also collected three pushcores of areas of sediment on top of the very flat sheet flows and found that the flows were covered by about four centimeters of mud. Previous research indicates that this part of the seafloor accumulates about one centimeter of sediment every 250 to 300 years. This suggests that the these flows could be 1,000-1,200 years old--quite a bit older than our initial estimates. If this age turns out to be correct, then the hydrothermal systems in this area may be relatively long-lived, since they should have started up at about the same time as the eruption and intrusion of lava into sills within the sediment.
We have somehow been keeping up with processing all the cores, but have not really had time for more than a cursory look at the glass fragments recovered from these cores. However, we already know that limu o Pele bubble fragments as large as 8 mm across have been recovered four kilometers south of the inferred vent. Samples from our earlier dives contained smaller particles at similar distances to the north and west. Thus, it is starting to look like the main current direction during the eruption may have been from north to south. When all the samples are fully processed, we will be able to make maps of the size distributions of particles at various distances and determine how the particles were dispersed and how high into the water column they were lifted, as Lionel Wilson explained in yesterday’s posting.
Stephanie writes: One of the things that complicates work at sea is trying to figure out where we are. We know where the ship is by using GPS navigation, but the ROV Tiburon is somewhere below us at the end of a two-mile-long tether, and you can't get satellite positions at the bottom of the sea!
The most precise seafloor navigation is done by deploying a network of acoustic beacons called transponders on the ocean floor. In this case, the submersible has a receiver which times the acoustic waves from the beacons. Using the velocity of sound in water (about 1500 meters per second, but it varies a bit with depth, temperature, and salinity), these sound travel-times are converted to distances. Knowing the distance from the submersible to at least 3 beacons, you can calculate the vehicle's position by trilateration. (For each beacon, draw a circle centered at the beacon with a radius equal to the distance to the vehicle. The intersection of the circles is the location of the vehicle). This is the same way seismologists calculate the location of earthquakes from multiple seismic stations. Many people mistakenly call this process triangulation, but triangulation involves three ("tri") angles, like when you're hiking and you find your position by using your compass to find your angle to three mountain peaks. When you use three distances instead of three angles, the process is more accurately called trilateration. When you navigate using transponders on the seafloor, it's called long-baseline transponder navigation. However, before you can do this, you must spend several hours or more deploying transponders and determining their positions.
For ROV Tiburon, we use a quicker navigation method called short baseline navigation. This inverts the process described above--the transponders are on the ship instead of on the sea floor. On the seafloor, you can have baselines that are a few kilometers long. However, short-baseline navigation is limited by the length of the surface ship. Sound waves from the submersible to the ship are turned into distances and the submerisble's position is calculated relative to the ship's position. A second instrument records the ROV's speed and heading, which is used to refine the position information.
Even when the ROV has good navigation, we have another problem -- we often need to figure out where we are relative to maps, bathymetry, or sidescan images that had poorer navigation. In 1996, we collected sidescan sonar data of the area we are presently exploring. This generates something that looks like a photo of the seafloor using sound waves instead of light. Unfortunately, during our 1996 survey, we had problems with our sonar transponder equipment. For most of the survey we only knew the ship's position and had to estimate how far behind the ship the sidescan sonar instrument was being towed. With a 3 kilometer cable, such position estimates can be quite far off!
On today's dive, we drove over a lava flow that shows up in the 1996 sidescan as a very bright sound reflector and mapped its outline, where it met the surrounding sediment. We could do this because the basalt lava is very hard, so it sends back most of the sound energy from the sonar. Sediments, on the other hand, are softer and absorb the energy, so they show up as dark patches on the image. While we were diving today, we carefully determined position of the contact between the sediment and the basalt and found that it was about 100m away from where it had been mapped on the 1996 survey.