Lake Bonney Expedition 2008

Going through the launch checklist, Bill inspects the o-rings before sealing up the radio beacon battery compartment

Maciek returns to the Bot House with one of the obstacles we have been avoiding near the limno hut, an ablation stake that has ablated out of the ice. It may be an object like this that snagged us a few days ago

Our status at the end of Mission 5

Thursday
December 16, 2008

Today’s profiler mission, Mission 5, took us towards the lake shore south of the Bot House. Our aim was to fill in the gap between where we had gone on Mission 3 and our keep out zone. The keep out zone is the area around John’s limnology tent, where we know there are active experiments hanging down into the water. To avoid entangling the bot or disturbing the experiments with either the bot or the fiber, we’ve had to designate a wedge of the lake as a keep out. Unfortunately this wedge means that there are a few grid points that we won’t be able to get to, even though they are within our range.

We ran through our launch checklist and dropped the bot in the water. The weather was much nicer today and the mission went quickly and cleanly. There was some worry that, even with our conservative stand-off distance from the limno tent keep out area, we might snag something under the ice on our way back to the Bot House, but Leah reported that the fiber was feeding back in as it should and the bot came home without incident.

Later that day, Maciek returned to the Bot House carrying a long pipe. He had spent the afternoon working around the limno tent and had encountered one of the experiments that defined a part of our keep out zone, an ablation stake. It had ablated out of its spot in the ice and was lying on the surface, so he removed it. While there are still experiments to avoid in the keep out zone, it is nice to know that there is one less object in the lake to trip us up.

Today the sonde camera recorded this image, among others, of microbial growths on rocks at 10 meter’s depth

Our status at the end of Mission 4

Thursday
December 15, 2008

This morning was deceptively warm and sunny. As soon as we got out to the bot house and started mission four, wind kicked up and the temperature dropped. Both the team at mission control and the team out tracking the bot had a chilly day. We had a few nervous moments where the communications link between mission control and the bot dropped out due to a kink in the fiber as it spooled out of the bot house but otherwise it was a textbook mission. In terms of the number of points hit, mission four was our most productive so far, with 13 new data points. We ran parallel to the north shore of the lake and got some exciting photos of the lake bottom.

We pulled the bot back into the Bot House at the end of the mission and put it on charge. Originally our plan today had been to run two missions but the tracking team was chilled and tired and we decided that the day would just run too long by the time we had recharged the batteries so we headed off to dinner.

Thursday
December 14, 2008

Now that we’ve completed three successful sonde missions, we took today off. We have been working 12 to 14 hours days out here and now that things are going well, it is time to rest. We stayed up late last night watching a Bruce Lee flick, so most folks slept in a bit and then we all went to Lake Bonney Camp to lounge about in the Jamesway. Leah baked cookies, since there is an oven at the camp, and we each had the chance to take a shower of sorts by using a solar shower bag filled with water we heated on the stove.

Our smiling assistant Leah from the BFC has come to help us stay sane. Good luck!

Shilpa and Vickie plan the upcoming mission. Mission planning is the balancing act of connecting the dots to hit as many points as possible without running out of battery power

Our status at the end of Mission 3

Thursday
December 13, 2008

Today we had a couple of changes in crew. We got a new helper, Leah, from the Berg Field Center in McMurdo to keep us sane and Peter and Bart left in the afternoon for McMurdo. Since Bart had been the only person designing mission plans for the bot each day, he took some time in the morning to show Bill and Vickie how to design a mission plan and then check whether the mission length and number of sonde drop points planned was within the bounds of our battery power. After the coordinates for the mission plan were extracted from the planning map and Shilpa programmed them into the bot, we were ready to get in the water.

Today’s sonde mission took us to the southwest edge of the lake. We were able to get the bot much closer to the shore than we had predicted. We hit 9 points today, including the points we had to skip from yesterday’s truncated mission. We continue to rack up data.

Our status at the end of Mission 2

Thursday
December 12, 2008

We were back to work as usual this morning. We planned a mission the southwest end of the lake, around the same area we navigated to on our first long run. Launch went smoothly; we are really getting the hang of our routine here now. Bill and Vickie tracked the bot on the surface and the folks in mission control watched as the bot navigated from point to point. A little over halfway through the mission, though, mission control radioed a message to the tracking team, “We seem to have the fiber snagged on something. We’re going to stop here and dive a little, see if we can get the fiber free.”

In order to get as many points covered as possible per mission, the missions are planned to be long narrow loops, usually working up one row of grid points and down the next row. Somewhere in this loop, we had wrapped the fiber around something sticking out of the ice ceiling. Now, even though the bot was heading closer to home, the fiber in the bot house was still spooling out, instead of going slack to be pulled back in.

After the dive, mission control let the bot continue its planned route and hit one or two more grid points. When the fiber continued to feed out, even as the bot approached closer and closer to home, we decided to cut a few points off of the end of the mission and head straight home. Since we have a 1000 m long string of fiber, we had enough line to bring the bot home, even with the fiber wrapped on some mystery obstacle several hundreds of meters away.

With the bot safely home and the fiber optic still wrapped around something out in the lake, the debate began about how to proceed. Some people wanted to recharge the bot’s batteries and go back down and, using the upward and horizontal cameras, follow the fiber back to find whatever it was caught on. Others wanted to detach the fiber from the bot and just pull it through by hand. By pulling on both ends of the line, the outgoing and the incoming we could feel that the line was wrapped around something smooth because there was not much resistance to tugging in either direction. In the end we decided that while finding and identifying the snag object would be useful for future missions, it was too risky to take the bot back out to investigate while the line was tangled—it might just make the problem worse. Instead, we decided to pull it through. We taped a small piece of foam to the metal connector on the end of the fiber so that it, like the rest of the fiber would float on the ice ceiling. We tossed the connector into the water and began to pull on the other end of the line. Everyone watched with anticipation as Bill pulled the line and fed it onto the deck as it came up. We all wondered if the connector would get stuck in the snag point, should we have cut the connector off? After a few minutes of pulling line in, Bill reported that there was suddenly much less resistance in the line, the connector must have made it around the snag. It was several minutes more before the end of the line made it back to the bot house. What a relief to get out of that mess. Although we still don’t know what we got caught on—an old experiment or an imperfection in the ice surface--we had the foresight to mark the cable before pulling it through, so now we know how far away the snag was, and in what approximate direction. Even though we had to cut the mission short, we still got 7 data points today.

Bart and Chris work on focusing the sonde camera

A Quickbird satellite image taken recently shows our camp and Bot House on West Lobe Lake Bonney

Thursday
December 11, 2008

Our successful sonde mission yesterday had us convinced that we were over the hump and would just be running uneventful missions from here on out. Chris, Bart and Kristof took some time in the morning to change the lens and adjust the focus on the sonde camera, so that our images of the lake floor would be crisp.

Once they were done, we sent the bot out on its mission for today, nine sonde drops. At the first grid point, we began to lower the sonde but less than halfway down the drop the data coming from the altimeter turned to garbage. The altimeter sits near the bottom of the sonde and tells us how far the sonde is from the bottom as it is slowly lowered. For environmental and hardware reasons, we want to avoid letting the sonde bump into the lake bottom. Normally the sonde is lowered until the altimeter indicates that it is 1 meter above the bottom. Without good data from the altimeter it is difficult to know when to stop the sonde drop and there is really no way to continue the science mission from that point, so we aborted the mission and called the bot home. Strangely, the altimeter resumed normal operation somewhere on the trip home.

Back in the bot house we started to work on the problem. We pulled the bot out of the water and, thinking that the garbage data might have been caused by a poor communications connection somewhere between the altimeter and mission control, we started to check the altimeter’s connections in the electronics housing in the sonde. Everything seemed fine, and anyhow, the altimeter now seemed to be working. We spent some time dropping the sonde in and out of the water and correlating the altimeter readings to the encoder on the spooler. Coupled with a depth-under-keel measurement from the sonar units, this gives us an additional way to know how far the sonde has been spooled out and how much distance is left as it approaches the bottom. Assuming that we don’t see anymore problems from the altimeter, tomorrow we’ll be back to running missions.

John adjusts the transponder

Peter marks the location of our first data point!

Chris downloads our first lake data from the Seabird instruments at the end of the day

Vickie prepares the flags that will be used to mark the locations of the sonde casts

Our status at the end of the first profiler mission

Wednesday
December 10, 2008

Back in business with our repaired profiler, today we planned our first drop sonde mission. With Bill and Vickie waiting outside to track the bot from the ice surface with the loop antennas, the crew at mission control sent the bot under the ice on its first real data collection mission. We were done with all of our preliminary check-out tests and now its time to do some science. The idea of the drop sonde missions is this: We have made up an imaginary grid over the whole lake. The lines of grid are spaced apart by 100 meters in the x direction and 100 meters in the y direction. Every spot where two grid lines cross is a point where we want to get water chemistry measurements. To get these, the bot navigates at a depth of 1 meter below the ice ceiling to a grid point in its mission plan. When it reaches the point it floats up until the four feet at the top of the bot rest against the ice, a maneuver we call ice - picking. Once the bot is ice - picked, it is stable and begins to lower the sonde. The instruments on the sonde take measurements of the water chemistry and so on as the sonde reels out. The sonde stops lowering at a distance of 1 meter from the bottom, takes a photo of the bottom and then the sonde reels back up. While the bot is ice-picked and dropping the sonde, Bill and Vickie locate the bot’s position from the surface and mark the spot with a flag. At the end of the day, Maciek uses a GPS to get a coordinate for each flag. We can compare these coordinates showing where the vehicle has actually traveled to where the computer’s mission log says it has traveled. This is a double check for our navigation and assures Peter and John that we have an exact an accurate location for the instrument measurements we’ve taken.

For our first data run, everything went incredibly smoothly. Bill, Peter and Vickie tracked the bot on the surface and watched as the flags marking the grid points lined up in 100 meter intervals. The folks at mission control found that things were going so well that we still had enough battery power to do an extra point on the way home. This gave us a total of 8 sonde casts for the day, a very satisfying start to our scientific work at the lake.

On the bot’s way back to the melt hole after the sonde casts, Kristof took the opportunity to troubleshoot some problems he had been having with the USBL. The USBL is one of the bot’s navigation tools and there are two parts to it, the large transceiver is mounted to the vehicle and the 30 cm long transponder hangs down in the melt hole. It is kind of like a homing beacon: transponder sends out a signal, the transceiver picks up that signal and can then calculate the distance and bearing to the transponder. It feeds that bearing and distance to the bot and the bot uses it to help navigate home. For the past few days we had not been able to get it to work, the transceiver was not receiving the transponder’s signal, but after some experimentation today, we found the best depth for the transponder to hang at and now that system is working.

Blog entries from previous weeks can be found below:

Week 1 - Nov 26, 2008 - December 2, 2008

Week 2 - Dec 3, 2008 - December 9, 2008