Weather Station Installation
Finally time to head to the field! Our site is, fortuitously, not located in the rainy coastal mountains, but far inland towards the Argentinian pampas, and constitutes the headwaters of the Río Coyhaique that supplies the city’s water. The area is dominated by various forms of deforestation: the initial burning to make room for sheep when the area was first settled in the middle of the 20th century, and then recent cutting to provide firewood for the city of Coyhaique. The drive is pure magic, rolling hills of golden grass radiating from soil hummocks, the rare wetlands host white swans, pink flamingos. It is deep autumn and the lengas (nothofagus pumilio) glow orange in the understory while the ñires (nothofagus antarctica) crown the canopy with fiery flecks of red.
Day 1: We head into the deforested watershed and look for a site for the first weather station – we have two stations, one for this watershed and another for the native forest nearby. I pick a windswept slope with evidence of the early burn – charred trunks mourn the loss of the organic soil layer that has eroded down into to the stream channel below. This slope will undoubtedly represent the extreme of deforestation – high wind speeds, exposure to radiation, lots of bare soil evaporation. We spend a good part of the day digging a hole for the post, and installing sensors for rainfall, humidity, temperature, windspeed and direction, solar radiation, and soil temperature and moisture at four depths. Much left to do – we hike quickly over to two other sensors and download the data onto a small field computer in between rain episodes. Finally we head down to the stream while Brian, my adviser, works on his own biogeochemical experiments, and I measure flow. For the non-hydrologist, measuring flow is a non-trivial effort. It is very simple to measure the ‘stage’ or the level of the water at any given time in units of length. But in order to turn this into a volume, we need to multiply by the cross-sectional area of the stream. To complicate things, the cross-sectional area itself depends on the stage, so we need many measurements to create a rating curve that relates stage to discharge…So I stand over the stream for a half hour measuring the velocity of the stream (i.e. length per unit time) at 3 centimeter intervals (length), entering the depth of the stream each time (length) – three lengths make a volume, and the instrument does all the integration for us! Two and a half liters per second.
Day 2: Much more rain than day one. We spend the morning walking around the undisturbed watershed finding the right place for the second weather station – a quiet forest, every ridge revealing mossy wetlands, whispy tree beards floating in the breeze, lengas everywhere exploding in fall flames, balls of mistletoe hanging high in the canopy above. These forests are pure magic, I feel at once content, humbled by the misty rain dripping down through the canopy crown, running down the gnarled trunks, seeping from the mossy mats, aster carpets, winding its way into rills and gullies. I discover that my adviser and I have a fundamental difference in understanding of what we plan to instrument: the work I have previously done in ‘hillslope hydrology’ has been analyzing the behavior of a large, mostly homogenous hillslope that feeds a stream system below flowing perpendicular to the flow on the hillslope. But Brian’s idea is to instrument a hillslope that has an ephemeral stream running parallel to the hillslope down to the stream below – this is difficult from my perspective, as ephemeral streams on hillslopes create a large amount of ‘heterogeneity’ in the slope itself – the slope geometry becomes much more complicated, the slope is no longer ‘north-facing’ or ‘south-facing’. Brian is willing to let me run with my ideas, so I stick the weather station on the larger, uniform hillslope. We install the station in the rain, fingers starting to freeze…One more flow measurement, this time five and a half liters per second, almost twice as much as the other stream! Still small enough to step across.
Day 3: We need to build a fence at the first site to stop the sheep from chewing the soil temperature and moisture wires. Brian has an idea to make a friend of one of the wood cutters by paying him $40.000 chilean pesos for 20 fence posts (an absurd amount). When we come to pick up the fenceposts, we find that they are not posts as much as trees. Still, we load them into the truck, drive toward the field site, and carry each one about 1/4 mile to the instrumented hill slope. We quickly realize that the webbing we were going to use for the fence will shade the soil sensors, so we decide to add more 'rungs' to the fence to keep the sheep out instead. We construct a small shelf for the Davis console and data logger. Since these are consumer weather stations, the Davis data logger has very little storage, so at the current logging interval of 30 minutes, I'll need to come out to the field every 2-3 weeks to download the data. Brian is ordering a couple cheap tablets that will live with the console and log the data. At the end of the day, we find some 'indian bread' mushrooms (Cyttaria darwinii) living on the branches of a lenga -- I pick a few for dinner that evening. Needless to say, future trips to the field will include a good deal of mushroom harvesting.
Frozen ground: I initially thought seasonally frozen ground dynamics would not be as relevant in the Patagonia as in Colorado. After plotting air temperature time series for the Coyhaique Alto area, I'm not so sure...The soil sensors closer to the surface are showing temps of 36-37 °C, and it's only fall!
Needle ice - see this post for a detailed explanation.
Day 1: We head into the deforested watershed and look for a site for the first weather station – we have two stations, one for this watershed and another for the native forest nearby. I pick a windswept slope with evidence of the early burn – charred trunks mourn the loss of the organic soil layer that has eroded down into to the stream channel below. This slope will undoubtedly represent the extreme of deforestation – high wind speeds, exposure to radiation, lots of bare soil evaporation. We spend a good part of the day digging a hole for the post, and installing sensors for rainfall, humidity, temperature, windspeed and direction, solar radiation, and soil temperature and moisture at four depths. Much left to do – we hike quickly over to two other sensors and download the data onto a small field computer in between rain episodes. Finally we head down to the stream while Brian, my adviser, works on his own biogeochemical experiments, and I measure flow. For the non-hydrologist, measuring flow is a non-trivial effort. It is very simple to measure the ‘stage’ or the level of the water at any given time in units of length. But in order to turn this into a volume, we need to multiply by the cross-sectional area of the stream. To complicate things, the cross-sectional area itself depends on the stage, so we need many measurements to create a rating curve that relates stage to discharge…So I stand over the stream for a half hour measuring the velocity of the stream (i.e. length per unit time) at 3 centimeter intervals (length), entering the depth of the stream each time (length) – three lengths make a volume, and the instrument does all the integration for us! Two and a half liters per second.
Day 2: Much more rain than day one. We spend the morning walking around the undisturbed watershed finding the right place for the second weather station – a quiet forest, every ridge revealing mossy wetlands, whispy tree beards floating in the breeze, lengas everywhere exploding in fall flames, balls of mistletoe hanging high in the canopy above. These forests are pure magic, I feel at once content, humbled by the misty rain dripping down through the canopy crown, running down the gnarled trunks, seeping from the mossy mats, aster carpets, winding its way into rills and gullies. I discover that my adviser and I have a fundamental difference in understanding of what we plan to instrument: the work I have previously done in ‘hillslope hydrology’ has been analyzing the behavior of a large, mostly homogenous hillslope that feeds a stream system below flowing perpendicular to the flow on the hillslope. But Brian’s idea is to instrument a hillslope that has an ephemeral stream running parallel to the hillslope down to the stream below – this is difficult from my perspective, as ephemeral streams on hillslopes create a large amount of ‘heterogeneity’ in the slope itself – the slope geometry becomes much more complicated, the slope is no longer ‘north-facing’ or ‘south-facing’. Brian is willing to let me run with my ideas, so I stick the weather station on the larger, uniform hillslope. We install the station in the rain, fingers starting to freeze…One more flow measurement, this time five and a half liters per second, almost twice as much as the other stream! Still small enough to step across.
Day 3: We need to build a fence at the first site to stop the sheep from chewing the soil temperature and moisture wires. Brian has an idea to make a friend of one of the wood cutters by paying him $40.000 chilean pesos for 20 fence posts (an absurd amount). When we come to pick up the fenceposts, we find that they are not posts as much as trees. Still, we load them into the truck, drive toward the field site, and carry each one about 1/4 mile to the instrumented hill slope. We quickly realize that the webbing we were going to use for the fence will shade the soil sensors, so we decide to add more 'rungs' to the fence to keep the sheep out instead. We construct a small shelf for the Davis console and data logger. Since these are consumer weather stations, the Davis data logger has very little storage, so at the current logging interval of 30 minutes, I'll need to come out to the field every 2-3 weeks to download the data. Brian is ordering a couple cheap tablets that will live with the console and log the data. At the end of the day, we find some 'indian bread' mushrooms (Cyttaria darwinii) living on the branches of a lenga -- I pick a few for dinner that evening. Needless to say, future trips to the field will include a good deal of mushroom harvesting.
Frozen ground: I initially thought seasonally frozen ground dynamics would not be as relevant in the Patagonia as in Colorado. After plotting air temperature time series for the Coyhaique Alto area, I'm not so sure...The soil sensors closer to the surface are showing temps of 36-37 °C, and it's only fall!
Deforested Slope - sheep fence in construction.
Deforested slope - the Davis console and data logger lives in a tree 100m from the station.
Forested slope - no sheep fence necessary here, but posts are installed to remind us not to step on the soil sensors.
Forested slope -- Davis console and data logger lives in a tree 5m from the station.
Forested watershed in the background.
Air temperatures at Coyhaique Alto. Long-term datasets! Woohoo!
Comments
Post a Comment