Baseline

The Projection

Chapter 4: Baseline

Carl's plane left at seven, the engine breaking the morning silence with the particular violence of a radial starting cold, the propeller turning once, twice, catching, the exhaust coughing blue smoke that drifted across the gravel bar and dissipated in the still air. The plane rolled downstream, turned into the light wind from the north, and accelerated, the wheels bouncing once on the gravel before the wings took the load and the Beaver climbed away over the river, banking left, heading south, the engine noise diminishing until it was absorbed by the sound of the water and the wind and the vast, implacable quiet of the valley.

Silas watched until the plane was a speck against the mountains. Then it was gone.

They were alone.

The aloneness was immediate and complete and had a weight that Silas recognized from thirty-eight years of fieldwork. It was the weight of distance — the distance from the nearest road, the nearest town, the nearest person who was not Jin Park. The distance could be measured: two hundred miles to Fairbanks, eighty miles to the village of Wiseman on the Dalton Highway, fifty miles to Bettles, which was the nearest place with a runway long enough for a wheeled aircraft and a satellite phone connection that worked reliably. In an emergency, Carl could be back in four hours if the weather allowed. If the weather did not allow — and in the Brooks Range, the weather frequently did not allow — they were on their own, and on their own meant the satellite phone and the InReach and the first aid kit and whatever competence they carried in their hands and their heads.

This was the condition of the work. Silas did not romanticize it. He noted it, the way he noted the weather (clear, fifty-two degrees, wind northwest at five knots) and the river level (high, running fast, the gravel bar submerged at the downstream end) and the condition of the camp (secure, food hung, equipment stowed). These were the observations that preceded the work, the reconnaissance that made the work possible, and he recorded them in the field notebook, on the first page, in the small, precise handwriting that he had developed over decades of writing in rain and cold and wind, a handwriting that was not beautiful but was legible, which was the only aesthetic criterion that mattered in a field notebook.

Rite in the Rain notebook number 373. Green cover. Waterproof pages. He wrote the date: June 8. He wrote the location: Koyukuk River, unnamed tributary, gravel bar, river left, approximately 67.42 N, 150.18 W. He would get the precise coordinates from the GPS later. For now, the approximate position was sufficient to identify the place.

"Ready," Jin said. He was standing by the GPS equipment, the R10 receiver mounted on a carbon-fiber pole, the data collector in his hand, the antenna blinking green to indicate satellite lock. He had set up the equipment quickly and correctly, and Silas noted this without comment, because competence was expected and praising expected competence was patronizing.

"We start with the baseline," Silas said. "Two points, as far apart as we can get on this bar, with a clear line of sight between them. The distance between the points is the baseline. Everything else is measured relative to it."

"I know what a baseline is."

"Then show me."

Jin looked at the gravel bar. It ran roughly north-south, following the river's bend, approximately three hundred meters long and fifty meters wide at the widest point. The willows crowded the upstream end. The downstream end was partly submerged. The usable area was roughly two hundred meters of open gravel, flat, clear.

"North end and south end," Jin said. "That gives us about two hundred meters."

"Two hundred and what."

"I'd have to measure it."

"Then measure it."

Jin walked to the north end of the bar, planted the survey pole, and stood there while the GPS collected data. After five minutes he walked to the south end and repeated the process. Then he looked at the data collector.

"Two hundred fourteen point seven meters," Jin said. "Plus or minus twelve millimeters."

"Write it down."

"It's in the data collector."

"Write it down in a notebook. If the data collector fails, if the battery dies, if you drop it in the river, the notebook survives. The notebook always survives."

Jin looked at him. Silas looked back. This was the first lesson and it was not about the baseline. It was about the relationship between the instrument and the record, between the ephemeral and the permanent, between the data that existed as electrons in a chip and the data that existed as graphite on waterproof paper.

Jin took out a notebook — not a Rite in the Rain, Silas noticed, but a standard field notebook with a blue cover that would turn to pulp in a rainstorm — and wrote down the distance.

"You'll need a better notebook," Silas said.

"This is what they gave me in the office."

Silas reached into his pack and pulled out one of the green Rite in the Rain notebooks and handed it to Jin. "Use this. Write everything in it. Every measurement, every observation, every bearing, every elevation. Date every page. Number every station. If it's worth measuring, it's worth writing down."

Jin took the notebook. He turned it over in his hands, feeling the weight of it, the texture of the waterproof pages. It was a small thing, a notebook, but it was the artifact around which the entire survey would be organized, the physical object that would carry the data from the field to the office to the map, and Silas had seen field notebooks from the nineteenth century, the original surveys of the American West, notebooks carried by men who measured with chains and compasses across landscapes that had no benchmarks, no roads, no maps, and those notebooks were still legible, still useful, the observations still valid a hundred and fifty years later. The data collector would be obsolete in ten years. The notebook would outlast them both.

They established the baseline. Silas set up the theodolite at the north end of the bar, centering it over the GPS point that Jin had marked with a wooden stake and a pink flagging ribbon. He leveled the instrument, adjusting the three leveling screws until the circular bubble was centered and the plate bubble confirmed it, and then he sighted on the stake at the south end of the bar, which Jin had also flagged. The horizontal circle read zero. He locked the circle and read the vertical angle: minus zero degrees, fourteen minutes, thirty seconds. The south end of the bar was slightly lower than the north end. The gravel bar sloped downhill toward the river.

"Record this," Silas said, and read out the values. Jin wrote them in the Rite in the Rain notebook in a handwriting that was smaller and neater than Silas's, the handwriting of a generation that had learned to write but had learned it as a supplementary skill, not a primary one.

"Now we need a backsight," Silas said. "A third point, off the baseline, that we can see from both ends. Something permanent. Not a flag. Not a stake. Something that will be here when we come back."

"The rock," Jin said. "The one you sighted on last night."

Silas looked at him. "Yes. The rock."

They spent the morning establishing the control network. Silas at the theodolite, measuring horizontal and vertical angles from the baseline to the reference rock and to four other points — a boulder on the riverbank, a dead spruce on the ridge, a prominent bend in the river, a patch of exposed bedrock on the opposite slope. Jin at the GPS, occupying each point for five minutes, recording the satellite-derived coordinates, logging the data in the collector and writing the positions in the notebook. They worked in a rhythm that was not yet natural but was becoming efficient, the rhythm of call and response — Silas calling out angles, Jin recording them; Jin calling out coordinates, Silas recording them.

By noon they had five control points in addition to the two baseline endpoints, and the positions derived from the theodolite angles agreed with the GPS positions to within eight centimeters, which was good for a first-order traverse in rough terrain and which told Silas that both the instruments and the operators were performing within acceptable limits.

They ate lunch on the gravel bar. Crackers, cheese, dried salami, a handful of trail mix. The sun was high and warm and the river glittered. Jin studied the data in the collector, scrolling through the positions, comparing them to the quad sheet that Silas had spread on a flat rock.

"The quad sheet says the tributary mouth is at 2,140 feet elevation," Jin said. "The GPS says 2,117 feet."

"The quad sheet was compiled from 1955 photography with a hundred-foot contour interval. The elevation could be off by fifty feet and still be within the stated accuracy."

"Twenty-three feet is within fifty."

"Yes."

"So the map was wrong."

"The map was imprecise. There's a difference. Wrong means the map lied. Imprecise means the map told you how much it didn't know. The contour interval is a statement of imprecision. It says: I can't distinguish elevations closer together than this. That's not a lie. That's an honest limitation."

Jin looked at the quad sheet. The contour lines crossed the tributary in wide, even curves, each line a hundred feet of elevation apart, and between the lines the map said nothing about the terrain — whether it was smooth or rough, gentle or steep, whether there were cliffs or terraces or benches that fell between the contour values and were therefore invisible, present in the landscape but absent from the map.

"What contour interval will you use," Jin said.

"Twenty feet."

"Why twenty."

"Because at 1:24,000 scale, twenty-foot contours are readable without crowding. Ten-foot contours would give more detail but they'd overlap on the steep sections and the map would be illegible. Forty-foot contours would lose too much. Twenty is the balance."

"The balance between what."

"Between what you show and what you hide. Between resolution and readability. Between the data you have and the map you want."

Jin considered this. "In GIS, we don't have to choose. We can set any contour interval. We can zoom in, zoom out. The data supports whatever resolution the user wants."

"The data supports it. The map doesn't."

"What's the difference."

"The map is the thing you look at and understand. The data is the thing you look at and query. Understanding and querying are different acts. When you query, you ask the data a question and the data gives you an answer. When you understand, you look at the map and the map gives you the terrain — not an answer, but the thing itself, represented. A good map doesn't answer questions. It makes questions unnecessary."

Jin was quiet. Silas was not sure he had communicated what he meant. He was not sure he could communicate what he meant, because the distinction was not intellectual but experiential, rooted in years of drawing maps by hand and seeing the terrain emerge from the contour lines the way a face emerged from a sketch, line by line, curve by curve, until the representation achieved a kind of presence that was not the presence of the real thing but was, in some way he could not articulate, sufficient.

Margot had articulated it. Margot had said: "The painting is not the mountain. The painting is what happens when the mountain passes through a human being and comes out on canvas." And the map, Silas thought, was what happened when the terrain passed through a human being and came out on vellum. The human being was the instrument. The most important instrument. The one that decided what mattered.

They went back to work in the afternoon. The plan was to walk upstream to the tributary mouth and begin the traverse — a series of theodolite stations extending up the drainage, each station connected to the previous one by measured angles and distances, the chain of stations forming a network from which the contour lines would be interpolated. But first they had to see the tributary, to walk its lower reach, to understand the terrain before they measured it.

The walk from the camp to the tributary mouth took twenty minutes over uneven ground — river gravel giving way to a bench of dried mud and willow scrub, then dropping to the tributary itself, which entered the Koyukuk through a braided delta of gravel and silt. The tributary was small, maybe fifteen feet wide at the mouth, running clear over a bed of grey cobbles. The water was cold — Silas dipped his hand in and estimated forty degrees — and the current was brisk, the water moving with the purposeful speed of snowmelt running downhill through a steep drainage.

They stood at the mouth and looked upstream. The tributary valley was narrow, flanked by ridges that rose steeply on both sides, the ridges covered with tundra on the lower slopes and bare rock on the upper slopes. The valley floor was perhaps a hundred meters wide, much of it occupied by the creek and its active floodplain, a zone of gravel and sand and occasional willows that shifted with every high-water event. Above the floodplain, on both sides, were benches of older alluvium, stable enough for walking, and above the benches the slopes began, rising at angles that Silas estimated at thirty to forty-five degrees, steep enough that the contour lines on the finished map would be close together, close enough to read the steepness without a legend.

"We'll run the traverse along the right bank," Silas said. "On the bench above the floodplain. Stations every two hundred meters, more if the terrain gets complicated."

"Define complicated."

"You'll know it when you see it. Complicated means the contour lines can't be interpolated from the station data alone. It means you need more stations, more measurements, more detail. Complicated means the terrain is doing something the instruments can't capture at the spacing you've chosen."

They walked upstream. The tributary bent to the north after two hundred meters, and the valley opened slightly, and the ridges pulled back, and the creek ran straight for a stretch through a flat area of willows and dwarf birch. Silas stopped and looked at the ground. He looked at the ridges. He looked at the creek.

"Station one here," he said.

"Why here."

"Because I can see the mouth from here and I can see the next bend from here. Line of sight in both directions. And the terrain is changing — the valley's opening up. The gradient's decreasing. The creek is spreading out. These are transitions, and transitions are where the map needs the most information."

He knelt and drove a wooden stake into the ground and tied a piece of pink flagging to it and wrote S-1 on the stake with a permanent marker. Station one. The first point in the traverse. The first measurement of the unnamed tributary.

Jin set up the GPS over the stake. The R10 blinked, acquiring satellites, computing a position from the signals of twenty-two spacecraft orbiting twelve thousand miles above the earth, each spacecraft carrying an atomic clock accurate to within one nanosecond, the clocks broadcasting time signals that the receiver used to calculate the distance from each satellite, four distances minimum, solving the equations of three-dimensional trilateration to determine a position on the surface of the earth: latitude, longitude, elevation. The mathematics were elegant. The engineering was extraordinary. The result was a set of numbers that told you where you were to within a few centimeters, anywhere on the planet, at any time.

Silas set up the theodolite next to the GPS. The Wild T2, fourteen pounds, optical, mechanical, manufactured in Switzerland in 1978 by Wild Heerbrugg, serial number 215847. He leveled the instrument. He sighted on the baseline stake at the camp, visible through a gap in the willows. He read the horizontal angle: one hundred seventy-three degrees, forty-two minutes, fifteen seconds. He turned to the reference rock on the ridge: two hundred twenty-eight degrees, eleven minutes, forty-five seconds. He read the vertical angles to both targets. He called the numbers to Jin, who wrote them in the notebook.

Then he did something Jin did not expect. He pulled out the Brunton compass and took a bearing to a point upstream, a distinctive rock on the creek bank.

"What's that for," Jin said.

"Redundancy."

"You already have GPS and theodolite."

"And now I have a compass bearing. Three measurements, three methods, three chances to catch an error. If the GPS says the rock is at azimuth two-forty and the theodolite says two-forty-one and the compass says two-fifty-five, I know the compass has a problem — maybe the rock is magnetic, maybe I misread the declination. But if all three agree within their precision, I know the measurements are reliable."

"That seems like a lot of work for a point you already have."

Silas looked at him. "Every point is a point you already have. The question is whether you trust it. Trust comes from verification, not from precision. A GPS position accurate to eight millimeters that you haven't verified is less reliable than a compass bearing accurate to two degrees that you've checked against two other measurements."

Jin wrote the compass bearing in the notebook. He did not argue. Silas appreciated this. The young man was learning, and learning in the field was not the same as learning in the office or the classroom. In the field, the lessons were delivered by the terrain, by the instruments, by the weather, by the physical experience of standing in a creek valley with a fourteen-pound instrument and a notebook and trying to reduce the landscape to numbers that could be drawn as lines on paper. The field did not explain itself. It presented itself, and you either learned to read it or you did not.

They worked upstream through the afternoon, setting four more stations, measuring angles and distances and elevations, recording everything in the notebook and the data collector, the dual record that was Silas's insurance against the failure of either medium. At each station Jin set up the GPS and Silas set up the theodolite and they measured the same points and compared the results, and the results agreed to within the expected tolerances, which was reassuring but not surprising because the instruments were good and the operators were careful and the terrain, in the lower reach of the tributary, was cooperative — open sightlines, moderate gradient, stable ground.

At station five, three-quarters of a mile from the mouth, they stopped. The valley was narrowing again, the ridges closing in, the creek running faster through a stretch of boulders and small rapids. The light was still good — it was eight in the evening but the sun was high in the northwest, casting the long shadows that Silas read the way he read contour lines, as indicators of slope and aspect and terrain shape.

"We'll stop here for today," Silas said.

Jin nodded. He looked tired. The work was not physically demanding in the way that hiking or climbing was physically demanding, but it required a sustained attention that was its own kind of exertion. Every measurement required focus — setting up the instrument, leveling it, sighting on the target, reading the angle, calling the number, recording it, checking it. The focus was cumulative. After eight hours of it, the mind was worn the way a pencil point was worn, blunted by use, needing to be sharpened.

They walked back to camp along the creek bank, following the flagged stations, the pink ribbons marking the traverse like a trail of breadcrumbs through the willows. The camp was as they had left it. The food was still hung. The tent was still standing. The river was still running.

Silas built a fire from driftwood and they cooked dinner — rice and beans and canned chicken, simple and hot — and ate sitting on the gravel bar, facing the river, facing the long light.

"Can I ask you something," Jin said.

"Yes."

"Why do you do it by hand."

"Do what by hand."

"The map. Ruth said you draw the final map by hand, on vellum, with pens. She said you're the last cartographer in the Survey who does it that way."

"I might be."

"Why."

Silas looked at the river. The question was not impertinent. It was the question he would have asked if he were Jin's age and watching an old man use instruments that belonged in a museum.

"Because the hand makes decisions that the software doesn't," Silas said. "When I draw a contour line by hand, I'm interpolating between the measured points, and the interpolation is a judgment. The line curves this way or that way based on what I know about the terrain — not just the numbers, but the shape of the ground, the way the water flows, the geology, the vegetation. The software interpolates mathematically. It draws a smooth curve between the points using an algorithm, and the algorithm doesn't know the terrain. It knows the numbers. The numbers are not the terrain."

"But the software is consistent. It applies the same algorithm everywhere."

"Consistent and correct are not the same thing. The software is consistently wrong in places where the terrain does something the algorithm doesn't expect. A cliff. A terrace. A drainage that cuts through a ridge. The hand can adjust. The algorithm can't."

Jin was quiet for a moment. "What if the software improves. What if the algorithm learns the terrain."

"Then it won't need me. That's fine. The work is the work, regardless of who does it."

"Or what does it."

"Or what does it."

They sat in silence. The fire cracked and popped. The river ran. The mountains held the last of the light on their high ridges, the light moving slowly eastward as the sun tracked north, and the valley was in shadow but the peaks were lit, and the contrast between the dark valley floor and the bright peaks was the contrast between the measured and the unmeasured, the known and the unknown, the mapped and the unmapped territory that surrounded them on all sides and extended without interruption to the Arctic Ocean.

Silas took out the field notebook and began writing the day's notes. Date. Weather. Stations established: S-1 through S-5. Distances measured. Angles recorded. GPS positions logged. General observations: tributary gradient moderate in lower reach, approximately two percent, steepening above S-5. Valley width at mouth: approximately one hundred meters. Valley width at S-5: approximately sixty meters. Creek width: twelve to fifteen feet. Substrate: cobble and gravel, some boulders above S-4. Vegetation: willow and dwarf birch on floodplain, tundra on slopes. Wildlife observed: lesser yellowlegs, arctic ground squirrel, fresh bear scat (grizzly, approximately two days old) near S-3.

He wrote these observations because they were the context that the numbers could not provide, the qualitative data that would help him draw the map when he was back in Fairbanks, sitting at his drafting table, translating the field notebooks into contour lines on vellum. The numbers told him where the contour lines went. The observations told him what the landscape looked like between the lines. Both were necessary. Neither was sufficient.

He closed the notebook. The fire was low. Jin had gone to the tent. Silas sat alone on the gravel bar in the half-light of the Arctic summer night and listened to the river and thought about nothing, which was the luxury of the field, the gift that the landscape gave you if you stayed in it long enough — the gift of an empty mind, a mind cleared of the clutter of the human world, a mind that heard only the river and the wind and the occasional call of a bird and the vast, encompassing silence that was not silence but the sound of a place that had no use for human noise.

He thought about Margot. He had told himself he would not think about Margot in the field, that the field was the place where the work pushed everything else aside, where the measurements and the observations and the physical demands of the day left no room for grief. But the field was also the place where he was most himself, most reduced to his essential function, and his essential function had been, for thirty-five years, defined in part by its relation to Margot's essential function, and without that relation the function was incomplete, the way a baseline was incomplete without a backsight, the way a position was incomplete without a datum.

The fire went out. He crawled into the tent. Jin was asleep, breathing evenly, his face turned toward the tent wall. Silas lay in his sleeping bag and looked at the ceiling of the tent, which glowed with the ambient light of the midnight sun, and he listened to the river, which would run all night and all the next day and all the days after that, measured or unmeasured, mapped or unmapped, indifferent to the numbers in the notebook and the positions in the data collector and the contour lines that would eventually be drawn on vellum by a hand that was, despite its thirty-eight years of practice, still imperfect, still uncertain, still making judgments about the shape of the ground that the ground itself would neither confirm nor deny.