In the summer of 2011 I spent ten days on a glacier. My group consisted of a professor, his graduate student, two local guides and helpers, and three porters. We were there to measure the size and position of the glacier in order to understand how it moved. Many people don’t think glaciers move, but they do. They are the snails among rivers, flowing down the mountain a few meters a year. One has to measure the location of the glacier every year for at least ten years to be able to say something about its motion. By the time I joined them in 2011, the group had been visiting the same glacier every year at the end of the summer season for five years. That year, on the last day of our trip, we discovered that the data we had collected made no sense.
I first met the professor, and learned of these expeditions, at my uncle’s. I was still in middle school, but the professor evidently made quite an impression on me, for when I was wondering what to do during my first summer of college I immediately thought of him. I visited him at his office to ask if I could join his crew. It was a little room on the terrace of the institute building, filled with cigarette smoke despite the open windows. The professor was a big man, bearded, with burning eyes. He asked me how I could be of use. “I can code,” I said, “maybe I could help with the data analysis?”
When we reached the glacier I realized how foolish my answer had been. The Himalayas are the tallest mountains in the world, and they are also the youngest. They are the debris thrown up as India pushes into Asia, and so are very rocky and crumbly. When we first came within sight of our glacier, the Professor had to point it out to me. The ice was invisible under a blanket of rocks.
The team had come up with an smart way to measure the position of the glacier. They marked several large rocks that were firmly embedded in the ice, and measured the positions of these rocks every year. From the movement of the rocks they inferred the movement of the glacier. Most of the work on the trip was clambering out among the rocks on the glacier, finding the marked ones, and then surveying their positions with respect to a fixed point on the mountainside. There was no work for a programmer.
Our local guides were the best at finding the marked rocks. They didn’t need a map, or a GPS; they had measured the rocks last year, and so knew exactly where to find them this year. We were also lucky with the weather, and so had finished all the measurements before the final day. The others stretched out in the sun, or enjoyed a book in the shade. But now that we had new data, I realized I could finally be useful. I decided to see what our measurements looked like. I brought up the previous years’ data on the computer. The rocks along the edge of the glacier marched forward like an advancing army. I plotted the new positions of the rocks. The first point I plotted was nowhere near the edge of the glacier. Neither was the second.
This discovery ruined the mood of the camp. It was already nearing August, and soon the monsoon rains would make it impossible to come back and collect more data, even if we had the funds. At the rate of one set of data points a year, this was a huge loss. But it was strange that the readings made no sense. We had calibrated the instrument before setting off. We had measured known distances to check it was working. We had to investigate, but we only had six hours of combined battery life in all our laptops.
The professor feverishly tried to figure out the logistics of getting a new set of readings. The graduate student and two of the guides double-checked our instruments, but they were fine. Meanwhile, I plotted the rest of the data points. I had to plot them all before I noticed something. This year’s measurements of the rocks were not scattered on the map at random, as they would be if the surveying instruments had been broken. They formed the outline of a glacier, albeit the reflection of one that started at one end of the actual glacier and went off into the mountains. The points had somehow been mirrored. We had mixed up left and right. We did the calculations, figured out how to recover the points, plotted everything again to make sure, just before our six hours ran out.
That night, the porters returned from the base camp to help us dismantle and carry the camp back down. They brought with them some mutton and a local liquor. We sat around the fire, eating and drinking. The glacier, below us, continued its slow march to the sea.
Siddharth Krishna is a graduate student in Computer Science at New York University. He can (usually) tell left from right.