Repairing Tectonic GNSS in Bangladesh's Tea Region

The remainder of my fieldwork focuses on the GNSS (the general term for GPS) instruments in eastern Bangladesh to study the tectonics and earthquake hazard.

Mike Steckler
March 20, 2024

Part 4 of a series on tracking landscape changes in Bangladesh. Read part 1, part 2 and part 3 here.

Having finished the servicing and repairing of instruments in the coastal zone to look at land subsidence and sedimentation, the rest of my fieldwork focuses on the GNSS (the general term for GPS) instruments in eastern Bangladesh to study the tectonics and earthquake hazard. The day after we returned to Dhaka, Austin left to return home. The following day Lin Shen, my postdoc, and Ken Austin, a geodetic engineer, arrived. Lin is using satellite radar data to image the land deformation in the IndoBurma subduction zone. This is a chance for her to see the area up close and personal. Ken is from the EarthScope Consortium, formerly UNAVCO, which is funded by NSF to support geodetic research. I’ve been talking to Ken a lot over the last few weeks as he helped me get the GNSS receivers in the south back online. This is the first time that I have met him in person.

4.1GroupDU 1 650x356.jpeg
Our group at the Dhaka University Geology Department, founded the same year as Lamont. From left to right: Sanju, Lin, Ken, me, Babu.

After I spent a couple of days in Dhaka, the capital, in more meetings, we met Sanju Singha, who has been caring for the instruments at Dhaka University, to pack up all the equipment that we will need. Sanju is staying behind for some job-related examinations, but will join us in another day. Lin, Ken, Babu and I drove east across the Meghna River to Comilla, near the border with India.  

4.2Mainamata 650x366.jpeg
Lin during one of her many photos and selfies at the ruins of Mainamati.

On the way, we stopped at Mainamati, the large ruins of a 1,200 year old Buddhist monastery. On the Friday afternoon before Ramadan, it was crowded with visitors and school trips. Fighting the crowd, I managed to order the tickets in Bangla. Once inside, we became celebrities, with scores of people taking selfies with us. After we left at closing time, we continued to our hotel. Since we had a large late lunch on the way, we skipped dinner and crashed for the night. 

4.3RPURKen 650x366.jpeg
Ken programming the new receiver and modem at Raipur (RPUR) with the original antenna from 2007 on the right.

When one lithospheric plate subducts beneath another, the two plates stick to each other. The upper plate gets deformed until the stress becomes too high, and then it jumps back in an earthquake. Since here the two plates are converging at ~12 mm or less than half an inch per year, it can take a long time before it reaches the breaking point. However, we don’t know how long or when the last earthquake here was, so we can’t tell when a large earthquake might happen, days or centuries. All we know is that we can see the places deforming with our GNSS. We will be visiting seven sites over the next week, all in need of repair or reinstallation to better define the plate motion and how the plates are deforming. Our instruments can measure positions to ~2 mm in the horizontal and 6 mm in the vertical. That means that after years of observations, we can detect motions of 1 mm/year or better, which still amazes me.

4.4CHNR 650x366.jpeg
Inside the equipment box at Comilla University. Unvisited by us since 2016 and chewed up by rats, it was somehow still working.
4.5CHNRtea 650x366.jpeg
While working on the GNSS site at Comilla University, one of the employees brought us tea up on the roof.

Our first place to visit is Raipur, about 2.5 hours to the SW. We had a station here from 2007-2016 in a government guest house, but had to return the loaned receiver to UNAVCO. Since everything is gone but the antenna on the roof, we set up the equipment box and solar panel on the roof. When we got up to the roof, we saw that trees on every side had grown and blocked the view of the sky. While a signal still came through, we worried about multipath—the radar signal from the GNSS satellites—bouncing off the trees. I looked at my existing processed data and could see the scatter of the data increase through time until at the end of my data it became very bad. This is no longer a good site. However, since it is over seven years since we acquired data from here, even noisy data will provide some additional constraints. For a while. We talked over whether to install here, but with no new site scouted and it being a weekend, and not having everything we need for a new site, we decided to go ahead and reinstall it. I may well move the equipment to a new site in a year or two.

4.6CHNRsolar 650x366.jpeg
Sanju and Ken (under the solar panel) set up the solar panel for COML near the antenna.

Lin and I installed the solar panel, while Ken set up the equipment box. After getting everything in place, the modem proved to be very temperamental. Ken alternated between trying two different ones, but eventually got it to work. Meanwhile, Babu and I went out looking for locks to secure the box and bolts to chain it to the roof. After cruising around, we finally found what we needed. By the time we were done it was almost 7:30pm. We didn’t get back until after 10 pm, glad that we had had a large late lunch. We had light snacks for dinner and Sanju arrived while we were eating.  

4.7Tea 650x366.jpeg
A view of the tea plantation, or tea garden, behind our cabin at the Tea Resort and Museum in Srimongal.

Next was a site at Comilla University, situated on top of the first anticline, or hill, made of the delta sediments that have been folded up by the subduction zone. It is only 30 meters (100 feet) high. The ones we will go to tonight in Sylhet are much larger. The hills in Sylhet are where most of the tea in Bangladesh is grown.

4.8ladderequip 366x650.jpeg
Carrying equipment up and down the sometimes flimsy ladders at our primary school sites could be challenging, especially the batteries. It usually involved someone on the roof, someone on the ladder and someone on the ground.

We collected data from this site for four years, from 2012-2016, until the funding ended. After the necessary greetings and cups of tea, we went to the roof. This site is plugged into the grid with a backup battery for power outages. The battery was gone and the power converter broken, the cables almost chewed through by the rat that scurried away when we entered, but amazingly the receiver was still working with over four years of data stored on it. I am thrilled. We tried to fix the power converter, but could not. Instead, we will set up a solar panel similar to the sites we will visit later.  This site also took longer than I imagined, so we finally reached our hotel in Sylhet, the most northeastern part of Bangladesh at midnight.

4.9cutting tree 366x650.jpeg
At JURI, trees blocking the sky view needed to be removed or trimmed. Here a local laborer with a machete climbed a tree to do it.

For the rest of the fieldwork, we are staying outside of Srimongal, the tea capital of Bangladesh at the Tea Resort and Museum. We have a three-bedroom cabin on the spacious grounds with a tea garden behind it. Since the Muslim holy month of Ramadan is starting soon, we are the only guests. We have five sites to service in four days, none more than a two-hour drive away. Southern Sylhet has a series of north-south hills where the delta sediments are folded and thrusted by the subduction zone. The flanks of the hills are covered in tea plantations. Our GNSS stations span the hills to measure how fast the land is being compressed by the subduction zone, building stress towards a possible major earthquake.

4.10 ladder 650x366.jpeg
At SSPS, the ladder snapped in half when I climbed it. Luckily I only sustained minor injuries.

With the last two long days, we decided to do what we hope will be the easiest site first. We slept in then went to the Ichachara primary school (ICPS), an hour and a half farther east. We downloaded the data, changed the batteries, and once again struggled with getting the modem connected. Still, we were able to finish everything by 3pm, including taking pictures with the teachers. However, we didn’t find a good restaurant until we drove all the way to Srimongal, so lunch was once again at 5pm. This gave us an evening to relax in the living room together. Only Sanju and Babu went out for a dinner as Babu starts his Ramadan fast tomorrow. The rest us, still pretty full from the late lunch, just snacked on chanuchur, roasted soybeans and protein bars.  

4.11abefore 650x366.jpeg
4.11bafter 650x366.jpeg
Before and after pictures of the water tank that we had to move so that it doesn’t block the view of the sky for the antenna.

For our next day, we headed to JURI, our farthest station in a medical clinic. Here the battery was completely dead. We changed it and the solar power controller. As usual, problems arose with the modem. This trip, it seems, the problem is always the modem. Eventually, Ken got the modem software upgraded and working. Once again at this site, trees are obscuring the sky view and we arranged to have some of them cut back.  

4.12biryani 366x650.jpeg
During Ramadan, many restaurants are closed during the day. We managed to find a place where we get take out biryani that we ate for lunch.

Afterwards, we visited Arman, who I worked with on two electromagnetic surveys in 2022, at his field office in Juri Town. He is working for BAPEX, the Bangladesh Petroleum Exploration company, helping to collect a 3D seismic survey in the area. Sanju and Babu know a number of his colleagues, so it was a bit of a reunion for them. We stayed until after sunset when everyone was able to break their Ramadan fast, then headed back for our dinner and some rest.

4.13 corrosion 650x366.jpeg
Ken scrapping the green corrosion off a GNSS receiver using local WD-40 and a toothbrush.

Now it was time to try two stations in the same day. It was an hour drive across the hills and a rice-filled valley to SSPS at another primary school. Here the receiver box was inside the headmaster’s office, since it was previously broken into on the roof. Lin helped change the batteries and Ken worked on upgrading the modem. I joined Sanju to deal with the situation on the roof. The school had installed a water tank right next to the antenna blocking out half the sky. The school agreed to move it. Sanju and the person moving it went up the ladder to the roof. As I followed, Sanju warned me to be careful that the ladder was weak. On my next step, the ladder broke in half sending me tumbling. I got a bit banged up, but luckily no serious injuries. The heel I landed on is still sore, and I banged up my hip and elbow, but I am recovering. After that, I watched from below as they relocated the water tank.

4.14monkeys 650x389.jpeg
Some of the monkeys that played

After three hours we were done and moved on to DCPS, a primary school on the same anticline (hill) as our hotel. As only Babu was fasting, the rest of ate our packed lunch at the school, which has a sturdy iron ladder. Then we went to work, changing rusted cables, the batteries, etc. Since it is only 10 minutes from where we are staying, Ken decided to take the modem back with us and return in the morning. After hours of work, including working with Jim Normandeau back at UNAVCO, the modem still would not take the upgrades it needed or recognize the SIM card. Eventually, Ken had to give up. No need to stop back here in the morning with no working modem. After he returns home, Ken will configure several new modems for me to get to Bangladesh and Sanju will swap them in.  

IMG 6903 1 523x650.jpeg
We pose in front of the famous “Tea Lady” statue that welcomes you to the tea growing region of Sylhet.

Our last stop was in Chanurughat (CHNR), a town in the middle of the flat valley between the hills. It is a college (high-school level in the US). The equipment box is in a first floor room. When we opened it, the receiver was working but showed no satellites being recorded. We could not get it to see the satellites. We tried a spare receiver we brought and that didn’t work. We went up to the roof and when we removed the antenna, we found it was cracked and filled with water. It is now a souvenir. After we replaced it with a spare, everything worked, except the modem, of course. Installed in 2012, the model was too old to run the current VPN used to download data, and none of our remaining modems were working. We concluded with a picnic lunch in a tea garden nearby.

IMG 6917 650x366.jpeg
Replacing the antenna at Chanurughat (CHNR) that turned out to be filled with water.

We had now finished all the GNSS work. All 16 stations were now recording data. Twelve of them are online and sending the data back to the States. Two sites unused since 2016 were resurrected as part of the trip. We collected a lot of data from the sites that were still running. We celebrated the end of the fieldwork with green coconuts and a final dinner out. This effort across Bangladesh will help us understand both the earthquake hazard in the east, and land subsidence in the south. The RSETs measuring elevation change, sedimentation and shallow subsidence had one of the twice yearly measurements. We collected ~230 sediment samples from many sites from up to 19 feet below the surface to help us understand the causes of the subsidence. This was a long trip with 12 people contributing, including the two drivers. Plus, there was the crew of the M/B Bawali. We drove the 4.5 hours back to Dhaka and Lin and Ken flew home in the evening. I stayed one extra day for yet more meetings and then followed.   

IMG 6921 650x366.jpeg
After completing our last station, we had a picnic in a tea garden with our boxed lunch from our hotel.