This week started out as a scorcher. Monday went without too much of a hassle when gathering data. We were only missing a chance to switch out the batteries for our Atmospheric Research Trailer for Environmental Monitoring and Interactive Science. In addition to gathering data, we underwent the exciting 3D printer training in Seeley G Mudd Library’s Makerspace! With the knowledge of using the 3D printers we were able to progress with making an adapter for the NOx Calibration Source to take 8oz N2O canisters in place of the 16oz canisters for calibration.
Our NOx Calibration Source is vital to confirming the trends and behavior of the NOx Monitor in order to help validate and better understand the data we gather at the Adam’s farm. Throughout the week, we ended up going through 5 different adapters trying to get the right shape with Tinkercad as our design engine. The first adapter we printed did not have tall enough walls to properly support the 8oz canister. With some quick adjustments, we were able to raise the walls for the second adapter. The third didn’t have thick enough walls to snuggly fit the canister. This knowledge pushed us forward to the fourth adapter, which I ended up giving the incorrect height. Being more careful since the last print, we made a working bottom piece! We finally tried to use the 8oz canister with the adapter’s help underneath the fume hood. Sadly, the shape of the puncture area for the canister on the top was too wide and let the gas escape. We hope to move forward with creating another solution for the top part of the canister with the experience we gained from creating the bottom piece.
Thursday held a challenge for us that we haven’t encountered yet. We had a large storm system coming through the Adam’s farm. Unfortunately, the batteries for ARTEMIS needed to be changed for a cloudy weekend, where the sun wouldn’t supply us with a self-sustaining amount of power for the solar panels on ARTEMIS. When we arrived on site, Nicki and I ran in with the 4 heavy batteries in tow and quickly changed them out. Immediately after the battery switch, we set up a structure securing the tarp over ARTEMIS so the instruments would not be ruined. Right after we secured the bungee cords, the second stretch of the storm came in heavy. We were able to breath with relief for a successful data gathering, calibrations, and battery switch for the day.
Week 7 held a lot of preparation for what will happen in the next two weeks. As Deanna will leave us to more independent work, Nicki and I started to focus more on coding and Raman Spectroscopy.
For a majority of the week, Deanna planned out the structure of the R language we are using to analyze the data from our NOx and Ozone Monitors, Kestrel, and TSI. So far, we have a lot of great correlation plots and are figuring out the optimal way to present our data. We are using the “OpenAir Manual” as a guide for understanding how to work the code for the plots and figures.
In addition to learning a new language, we finished our tutorial through using Raman Spectroscopy Instruments. With a tiny chip of silicon, a faux pearl, and real pearl, Nicki and I were able to see the differences the data we gathered and successfully gathered similar data to those from previous findings. These new skill sets will bring us farther in our data analysis.
We have had some pretty frustrating challenges this week, mostly with the instruments. The NOx cable connection is extremely weak – every time we so much as touch the NOx box, it powers off, making data download really finicky and personally stressful. At first the TSI wasn’t logging data for more than 24 hours, but once we solved that problem and Fry and I went out to tend to ARTEMIS on our own, it presented new, never-before-seen problems that we hope we have solved by now. We have made little progress in making Arduino CO2, noise, and dust sensors that log data, but that isn’t one of our highest priorities.
Another pretty serious drawback we encountered this week was learning that that the 16oz N2O canisters we use to calibrate the NOx box are discontinued worldwide. We discovered this when Deanna thought she ordered the correct size canisters, but when they came half the size, we investigated and learned this news. So now we have a new project up our sleeves: we will be 3D printing an adapter for the N2O canister sleeve, so that we can center and fit the smaller canister inside, and hopefully the puncture will work as usual. Though we do have this tentative solution to the problem, it directly affects us every ARTEMIS visit because we are now unable to calibrate the instrument until we 3D print the adapter. Although the NOx box is less likely to turn wonky on us, I am concerned that the longer it takes to calibrate the instrument again, the more potential for our data to become weird.
In other news, Rstudios has been the positive light in the week. We can now officially make plots with all the data, including the TSI particle data. Now the next challenge is to find appropriate color schemes for our complex correlation plots.
Here is an example of one of the plots we recently made: this shows TSI particle counts over the last couple of days, with each species noted by particle size in (μm).
The day of the Lawrence University Science Institute carried eventful show casing of our summer research. With bright eyed prospective students and a lot of work to be done, there was not a boring moment. Nicki, Deanna, and I met up with our prospective students interested in Chemistry at Lawrence. We hit the road to Adam’s farm around 9am, where ARTEMIS was collecting data. In the car, we talked about our research and the college searching experience with our morning student. Remembering how stressful it could, be was a real throwback in nostalgia for Nicki and me. We shared our experiences and journey of coming to Lawrence until we reached the site.
We started collecting data, while Deanna was explaining more of the projects and reasons behind collecting Ozone, NOx, Particulate Matter, and Weather data. (Sadly, I still cannot reach the Kestrel even with the new and taller ladder.) This involved concerns surrounding Frac Sand Mining and how ozone acts in the Stratosphere and Troposphere.
Later we taught our prospective student how to reconnect the tubing running from the Monitors and Calibration sources using wrenches. However, the day did not go on without a few bumps in the road as our NOx Monitor data transmission got interrupted and we had to wait another 30minutes for the data to resubmit to our laptops.
Getting back on campus led to another fun talk in the Chemistry department with all our prospective students over lunch. We introduced ourselves and shared exciting disaster stories regarding our equipment exploding, flying away, and fruits of our work being lost.
After refueling for the other half of the day, Nicki and I met up with our next two prospective students to go back out to the field at Adam’s Farm. Our second round of calibrating and gathering data was filled with sharing experiences at Lawrence and knowledge about ARTEMIS. We encountered a good period of waiting time while the calibration sources warmed up. During this time, we talked more about the possibilities in classes at Lawrence and gained a few grass braids I proudly donned for the rest of the day. The long wait time for calibrations took its toll as we only just made it back to campus on time by 4:30.
This past week held some major successes for us. We spent last Tuesday and Wednesday planning and constructing our particle sampling tower. It was really fun, I used a ton of tools that I’ve never used before in my life! (Learning life skills on the side). We were pretty worried about the actual installment of the tower though – we were just not confident that it would stay up, since the TSI box is about 50 lbs.
However, most of our worries were mitigated once we actually arrived on site on Thursday. We had planned to spend the entire day there, since we weren’t sure what kinds of obstacles would arise. But much to our luck, not many did! The stake went in the ground and we leveled it out without much of a fight. Daniel Martin, Lawrence’s stockroom supervisor, graciously came out to the site for a few hours to help us carry the tower and new, (heavy!) super tall ladder out 3/4 of a mile to ARTEMIS’s location. The tower went onto the stake quite easily, and we got 5 out of 6 bolts in to secure it.
I would say that the biggest challenge we faced that day was putting the tower anchors into the ground. The first one went into the ground easily! But the next two put up fights. Deanna wound up actually mallet-ing an old crowbar into the ground, since after 3 attempts with this one anchor it just refused to cooperate. We also wound up using some cinder blocks as a make-shift anchor to another line. Our security cables installed fine. We eventually moved ARTEMIS closer to the tower, because we had to connect the TSI to the power source, and the wires weren’t extremely long.
All in all it was a long and hard day, but everything wound up much easier than we were expecting. We ended the day with lemonade slushies as a personal reward!
The newest installment of ARTEMIS!
We had a couple of big successes this past week! First, Deanna wrote out a checklist of tasks for Fry and I to complete at the field. The test: to do it all on our own. And despite it taking about 3 hours, we did! A couple of hiccups here and there but we tended to ARTEMIS mostly on our own. Second, we saw our first NOx peak in the data. We are actually seeing stuff!
Third, a solution to how to compile all of our data came much faster than expected. Deanna coded her own way of compiling everything together using RStudios, so now we are one step closer in figuring out how we want to analyze all of the data we are collecting.
In other news, keep your eyes peeled for a branding logo for ARTEMIS. It may be coming soon!
The above image shows one of our many insect friends found at the farm.
Gathering Ozone, NOx, and Kestrel data has become more efficient with Deanna’s checklist for ARTEMIS operations. We finished in a record time of 3 hours and with more independence than before! Luckily we only encountered Sunny and partially cloudy weather for our data gathering so far.
Data organizing has gone more smoothly within the two days of gathering. We are still working on finding an optimal way to display all the data sets that we have relating to weather, Ozone levels, and NOx amounts. This will be a continuous process as we put our focus towards R.
For our background research on general principles of Atmospheric Chemistry, we’re diving into feedback loops, sinks, and other mechanisms concerning CH4, NOx, HOx, CO, and Ozone.