Notes from the Field

Doing Science in a Snowpit

January 27th, 2012 by Maria-Jose Viñas

By Ludovic Brucker

The main objective of deep field traverses like ours is to make in-situ measurements and collect samples, which obviously cannot be done from air or space. So, during the 2010-11 and 2011-12 SEAT field seasons, we have been taking in-situ snow and radar measurements.

We typically had two different types of days during our traverse (both equally enjoyable!): traveling days, where we would move to a new site and set up camp, and days where we would collect snow measurements and drill a 20-meter deep snow core near our camp.

(Actually, we had a third type of day where we’d spend hours and hours shoveling during snow storms… but that’s a different story!)

My duties during the days we worked around our camp included digging a 2-meter deep snowpit and recording various measurements. When traveling on snowmobiles, or even while standing on the surface of the ice, one has no idea of what the snow lying beneath looks like. Amazingly, I often got a clear idea of the snow layering when burying anchors for setting up our tents. This provided a first quick overview of what it would look like the pit on the following morning.

Typically, my days started at 5:30 AM, when I would shovel snow in a cooking pot to get warm water for the rest of the crew, who were supposed to get ready in the following hour. The next step was to find the best place to dig the snowpit. Usually, a good place consisted of a uniform zone ranging several square meters and representative of the surrounding area. It also had to be free of sastrugi (small snow ridges created by the wind). Additionally, I had to take into account the sun’s position and how it was going to move across the sky in the following hours: it takes some time to dig the pit and collect measurements, and we absolutely didn’t want that the sun ended up warming the snowpit face where we’d be recording our measurements. Once I had decided on the location (usually a matter of seconds), I’d start digging.

After a good first hour of shoveling, I would have breakfast. Then I’d come back to excavate down to about 2 meters. The deeper the snow, the harder the layers got. I would usually shovel 6-7 cubic meters (211-247 cubic feet) of snow. The snow weighted about 300-400 kg (770-880 pounds) per cubic meter. I’ll let you do the math, so you understand that my colleagues’ help was always warmly welcomed!

Once the pit was dug, I would try to smooth the snow surface as much as possible, covering an area wide enough to simultaneously record the following measurements:

1) Infrared photography

We put an infrared filter in front of the sensor of a standard reflex camera to record the snow cover stratigraphy (that is, the stacks of snow layers). This modification allowed us to take pictures at infrared wavelength (700-1100 nm), which is more sensitive to snow properties (such as grain size) than visible light is.

So, by simply taking a couple of photos, we were able to identify the various snow layers. We also used reference panels to calibrate our photographs and to get an idea of the vertical snow grain size variation with depth.

2) Visual stratigraphy

After taking infrared pictures, I would visually identify the snow layering simply by looking at the snow wall. Each time I saw a change in grain size, snow structure, and so on, I would mark a new layer. This process is less accurate than using infrared pictures, because it can be visually challenging to identify smooth vertical variations from one layer to another. However, I’d normally be able to distinguish the main features. I would then take snow grain size and hardness measurements for each visually identified layer.

3) Snow temperature

We measured snow temperature at every 10-cm interval, from the surface to the bottom of the pit. Most of the time, surface temperatures were about -10oC (14oF), and the lowest temperatures (found 2 meters below the surface) ranged between -25 and -29oC (-13 and 20oF).

4) Thermal conductivity

Thermal conductivity values tell us about how efficiently heat propagates through the snow cover. This way we can understand how quickly air temperature variations can travel down the snow. Estimating heat transfer through the snow cover is important in modeling the energy exchange between the snow and the atmosphere, among other characteristics.

We made our measurements using three needle probes: the one in the center continuously measures snow temperature, while the two on the sides slightly warm the snow surrounding them. The idea is to measure how fast and with which intensity the heat wave reaches the temperature-recording needle.

5) Snow density

Another important characteristic of snow is its density, or mass for a given volume. We made those measurements at two vertical resolutions (2 and 10 cm) using density cutters (thin metalic shapes that I inserted in the snow wall to extract samples). I did know that hard snow layers exist in Antarctica, but I was surprised to find out how difficult it was to insert a thin sharp metal piece into these layers! We bagged and preserved the 2-cm-thick samples (they will be shipped to the BYU lab in Utah for further analysis) and we weighted the 10-cm-thick samples directly in the pit.

Our route went across the West Antarctic Ice Divide, a topographic feature formed by a large-scale weather system (an atmospheric depression that comes from the ocean and creates snow storms inland).

We traveled across this divide, which presents very different snow accumulation scenarios on each of its sides. On Byrd’s side, where we started and ended the traverse, snow accumulation is low. But on the other side of the divide, there’s more snow accumulating each year.

Each camp we worked at presented different snowpacks, though they shared specific distinguishable layers at given depths. For instance, there was a layer with large snow grains at 0.92 m depth – this layer was visible throughout hundreds of kilometers.

In summary, during the 2011 SEAT traverse, we dug seven snowpits and collected all the above measurements down to 2 m deep. Now, it’s time to get in front of the computer to analyze and understand these incredible data sets!

Camp Life

January 24th, 2012 by Maria-Jose Viñas

By Michelle Koutnik

Each of our six different camping sites consisted of one cook tent, four sleeping tents, and a bathroom area (more on that later). The cook tent was a “Scott” tent, which is an enduring style and named for the polar explorer Robert Scott. It was a tight space for five people but we were able to crawl in and then sit around together with one person managing the two-burner propane stove in the corner. The Scott style tent was much sturdier than our mountain-style sleeping tents, but it was also more time consuming to put up and heavier to transport. Since we were moving nearly every other day, we wanted to keep our camp as simple and as light as possible. The Scott tent was a necessary addition for cooking, but also as a reliable shelter in the two strong storms we experienced during the traverse.

It took a few hours to set up camp, and a similar amount of time to break it down and repack the gear on the sleds. Randy, Jessica, and I were responsible for breaking down most of the camp and for setting up some of the camp on our own while Clement and Ludo collected radar data on traverse days. Once camp was set up at a new site, we could start melting snow for water and start cooking dinner.

On a traverse day, our camp life included the take down and setup of camp, plus collecting radar data. On an ice-coring day, our camp life included digging and sampling the snow pit, drilling the ice core, and eating a hot lunch. We ate well out there! Our standard meal plan included tortellini with pesto, spaghetti with meatballs, sausage with rice and vegetables, and burritos.  Sausages were popular and made their way into many meals. Special nights featured hamburgers with tater tots, and on Christmas we cooked scallops with rice and vegetables. Hot lunch on ice-coring days featured cheesy bagels fried with butter in our cast-iron pan. Since all the cheese and bread was always frozen, it required frying it all together before we could eat it – but it tasted great this way! Stormy-day food was simplified but did afford us the time to enjoy pancakes (two storms, two pancake breakfasts).

The storms also affected our standard bathroom situation – usually a snow pit and tarp configuration. Ludo took the high winds and blowing snow as a challenge and created a snow-brick bathroom, which we tried our hardest to maintain from drifts!

Maintaining camp and completing the science goals was a big job but we enjoyed being out on the ice sheet. For example, Christmas was a wonderful day to share together because a big storm started to clear, and we had lovely gifts to exchange, two small Christmas trees, and a very nice meal. Overall, the camp life was comfortable and enjoyable and with such a hard-working team, we were efficient at all the tasks necessary to make the camp run and be successful with the science.

Radar days on the West Antarctic Ice Sheet

January 23rd, 2012 by Maria-Jose Viñas

By Clement Miege

Hi there! After more than three weeks spent in the field, our team is very happy to be finally back, with many memories of the traverse. This year has been a very intense experience and I would like to tell you a little more about this expedition. I will focus on the days we were doing radar surveys. Indeed, two different studies were set up during this field work. The first one was during the traverse days, where we did a radar survey from one camp to another with a range of 50-90 kilometers (31-56 miles) between the camps. On the other days the surveys were smaller, set up around each camp; they consisted of a 10-km bowtie and a 280- by 280-meter (918- by 918-ft) grid to get a better idea of the spatial variability of snow layer depth surrounding each core site. These grid surveys will show us how representative the ice core is compared to the surrounding area and help answer the question of whether we would have gotten a different result if we had drilled our ice core a few paces this way or a few paces that way. We were also looking more in detail at the layering in a 2-meter (6.56-ft) snow pit; to do that, we were using a metal plate at different depths of the snow pit.

Here is a picture showing the two radars that operated simultaneously on the sled:

Both radars looked at the first 20 meters of the snow pack, sending electromagnetic waves into the snow. The snow radar (the big green horns in the picture) operates in a frequency range of 2-8 GHz. The other radar (the smaller brown horns) is the ku-band radar, sweeping between 12 to 18 GHz. The lower the frequency, the deeper the radars look. We have both radars for some redundancy in the system, imaging the snow/firn closest to the surface, which we core twice, and then the snow radar peers deeper than the cores to provide literally a deeper look. In addition to the 2 radars, we needed to know the elevation we were at, which is important for comparison with airborne data and also for modeling precipitation and temperatures. For that, we simultaneously collected GPS coordinates, which gave us our exact latitude, longitude and elevation every 5 seconds. The GPS antenna was on the very top of the sled, about 2.5 meters (8.2 feet) high.

On this sled, beside the radars sitting in the red box, we had packed: a blue bag, two black duffel bags and an orange bag. The blue bag was a survival bag, with all the gear needed in case we would have been caught in a storm; it included a tent, a stove, a shovel, and some food. The two black bags contained Ludo’s and my sleep kits, and the orange bag had some extra food, in case we were stuck for couple days. For the snow mobile, we had a repair kit under the seat to troubleshoot a possible failure. We were carrying two extra cans of gas as well. With this set-up, we were able to get a pretty light sled that was still comfortable and had all the gear we’d need in case of a storm or other circumstance separated us from the rest of the team, who were carrying all the camp gear. Fortunately, this never happened!

But now, let’s go back to the survey. The travelling days were the ones where the team was the most vulnerable: while we were traveling we had no shelters or camp set up to get back to if anything went wrong because we had to break down the entire camp in the morning, pack all gear on the sleds and ride with the snowmobiles pulling the sleds to the next camp. After driving between 50-90 km (31-56 mi), we would build our new camp, usually that same day in the afternoon. We had a total of seven travel days, covering a distance of about 500 km (310 mi).

During travel days, the radar team left camp first, mostly because we had to drive at low speed to ensure the radar’s safety. Indeed, on the days we ran into large sastrugi (small ridges of hard snow), we drove at less than 10 km/h (6.2 mph). These sastrugi made our travel a little bit more difficult. Toward the end of the day, the team responsible for breaking down camp would leave the radar team and go 20 km (12 mi) ahead, to start setting up the new camp.

The days we were at camp, we did some small radar surveys: a bowtie and a grid around the core site. To help us drive the snowmobile straight for the grids, we set up flags to visualize the corners and sides. In the deep field, everything is white and flat, so it is hard to maintain a nice bearing just by following the GPS for such a small grid.

The last side survey with the radar was done in the snow pit. After analyzing the snow pit and picking up snow samples for further laboratory analysis, we used the pit to calibrate the radars and look at a detailed snow layering for the top 2 meters.

To conclude, I want to say that this traverse was definitely an amazing experience and I was happy to share such a good time with the team. I am already excited to start looking at the radar data in detail to see what we can learn of the past few decades of snow accumulation in this part of the West Antarctic Ice Sheet.

GCPEx: GPM Cold-season Precipitation Experiment

January 20th, 2012 by Kevin Ward

GCPEx Logo

The GPM (Global Precipitation Measurement) Cold-season Precipitation Experiment (GCPEx) will be conducted in cooperation with Environment Canada in Ontario, Canada from January 17th to February 29th, 2012. The overarching goal of GCPEx is to characterize the ability of multi-frequency active and passive microwave sensors to detect and estimate falling snow through the collection of microphysical property data, associated remote sensing observations, and coordinated model simulations of falling snow. Through collection of these unique datasets, GCPEx will seek to improve the GPM snowfall retrieval algorithms.

The GCPEx experiment will use instrumented aircraft (NASA DC-8, NASA-funded University of North Dakota Cessna Citation, and Canadian National Research Council Convair 580) for flights over heavily-instrumented ground sites located in and around the Environment Centre for Atmospheric Research Experiments (CARE) located in Egbert, Ontario. The DC-8 aircraft will fly high above clouds and precipitation with instruments similar to those on the GPM Core satellite. The Citation and C580 aircraft will fly through snowing clouds to measure snowflake properties in situ. Ground-based equipment such as radars and surface particle and snow water equivalent measurement instrumentation will connect airborne measurements of snowfall to what is measured at the ground. Data from the experiment will be used to develop and validate snow and frozen precipitation retrieval algorithms used in the generation of data products for GPM, CloudSat and future polar precipitation missions planned by the European Union.

For more information about GCPEx:

GCPEx Overview

GCPEx Campaign Blog

Ground Validation Image Gallery (recent images from GCPEx)

You can also follow this campaign and other NASA precipitation measurement missions on Facebook

Synopsis of the Traverse

January 19th, 2012 by Maria-Jose Viñas

By Michelle Koutnik

Now for a recap of our adventure! We arrived in Christchurch on November 19 and returned there on January 5. We spent 17 days in McMurdo before leaving to Byrd camp on December 7. It took only a few days to prepare for the traverse and we left Byrd camp on December 10.

The traverse lasted 18 days, with the longest time spent at camps 4 and 5 due to the storm delays. Otherwise, we moved fast! We spent an extra day at camp 3 to drill a second ice core, but by the end of the trip we were such fast drillers that we were able to drill two cores in one day at camp 6. We drilled ice cores at nine different sites (including Byrd Station), dug and sampled 6 snow pits, and collected more than 500 km (310 mi) of radar data.

Back at Byrd, we broke down our gear and with the help of the Byrd cargo handler we had it all packed on palettes in one day. Ludo and Jessica went by Twin Otter to pick up the ice cores on the day after we returned to Byrd so that soon after the traverse ended we were finished with almost all our work! In 18 days we finished all of our science, but to achieve these goals it took near seven weeks of travel and preparation – it is not easy to do work in Antarctica!

After the traverse was finished, we could not get a flight out of Byrd Camp back to McMurdo station for a few days so we enjoyed our time with the Byrd Camp crew and rang in the new year with a gorgeous dinner and dancing with the whole camp. Then we had a fast two-day turnaround from McMurdo to Christchurch. We all worked to clean and return all the gear we used in the field and ship all of the science equipment back to the U.S. It was very satisfying to complete all of our goals and finish on time despite weather and flight delays during the season. Great work, team!!