Greenhouse Gas Research: Collecting Data

With the second summer session beginning to wind down, greenhouse gas data collection is well under way!  Since my last blog post, I’ve explored greenhouse gas emissions on new parts of Lake Matoaka and began sampling parts of the Crim Dell.  While an infinite amount of data could be collected in order to precisely quantify emissions, the next step is to use the data I’ve collected this summer to calculate an approximation of emissions from these bodies of water.

Morning paddle to a sampling location on the other side of Lake Matoaka

Morning paddle to a sampling location on the west side of Lake Matoaka


In order to get an idea of emissions at different depths and distances to shore,  I took samples from five different locations on Lake Matoaka.  I sampled each location at 9:00 AM in order to standardize the time of day emissions were measured, as previous data on the lake showed time of day to be influential.  However, it was impossible to control for factors like cloud cover, temperature, wind, and rain.  This means that the weather on the day I sampled a particular location may have had a large influence on the flux I measured, and that variation between locations is partially attributable to variation in weather between sampling days.

Still, there are some trends.  The Keck Dock and Boathouse Dock had the highest emissions and were the shallowest sampling locations, right along the edge of the lake, where they likely receive lots of nutrients and organic matter from land and may have higher rates of decomposition than sites in the center of the lake.  The Algae Dock, Deepest Point, and West Arm were all sites in more open water, where sunlight (particularly on clear days!) appeared to stimulate photosynthesis and sometimes led to an uptake of CO2.

Graph of average emissions on Matoaka and map of sampling locations

Average emissions and map of Lake Matoaka sampling locations


I began Crim Dell sampling by deploying the chambers from the spillway side.   These first samples immediately showed much higher levels of carbon dioxide and methane than most Lake Matoaka measurements.  The samples from the spillway also hinted at a slight positive flux of nitrous oxide, the elusive third greenhouse gas I am studying this summer.  While usually only detected at minute concentrations, nitrous oxide has a global warming potential that is 298 times that of carbon dioxide.  This means that any given amount of nitrous oxide will impact the climate 298 times as much as the same number of moles of carbon dioxide, so it’s important to keep tabs on this sneaky greenhouse gas.

Sampling the Crim Dell spillway

Sampling the Crim Dell spillway

The back of the Crim Dell has proven to be a different story entirely, surpassing all previous data by a wide margin in all three measured greenhouse gases.  This little body of water receives frequent flushes of storm water from a pipe that drains parts of old campus, and is separated from the rest of the Crim Dell by a sedimentation fence.  I sampled this part of the Crim Dell via canoe, twice around 9:00 AM and once at 11:00 PM.

I also began calculating emissions in terms of CO2 equivalents.  Since every greenhouse gas has a different heat-trapping capacity, transforming emissions data into CO2 equivalents shows the relative impact each of the gases is having on the climate, based on their emissions from a specific site.  In the Crim Dell, methane seems to be the dominant greenhouse gas.  Even though it is being emitted at slightly lower quantities than CO2, its global warming potential is 25x higher.  While nitrous oxide has an even higher warming potential (298x) than methane, so far it’s only been observed at very low fluxes, giving methane the highest relative impact.

There’s still a lot to explore in terms of Crim Dell emissions, and two other storm water ponds on campus that have not been sampled yet!  Then it’s time to crunch some numbers and get a rough estimate of  total emissions for our campus ponds.


  1. rjschwartz says:

    Your research is really interesting and it sounds like it’s been going super well this summer! I love that you’re doing the project on campus; it’s really cool to read this and have it be so relatable because we all actually live here. I was wondering what you’re planning on doing with the data after you finish your project. Do you hope that it can be used to make a plan to lower emissions? Is there anything that the community could do to potentially lower the emissions? Good luck with the rest of your project!

  2. carolina says:

    Thanks! Good question. I hope to keep gathering data throughout the school year to get a better picture of what annual emissions we are seeing from bodies of water on campus, and how seasonal factors affect emissions. While my project hasn’t been focusing on an emission-reduction plan, there are a few options that might lower emissions. The first would be to reduce the number of these storm water ponds being built. Of course, this would mean a need for some other means of controlling storm water, like permeable pavement or rain gardens. The second would be to minimize factors that lead to high greenhouse gas emissions, such as limiting nutrient pollution by using less fertilizer. As always, more research is needed to understand what factors play a role in the greenhouse gas emissions from storm water ponds, and what actions might be able to reduce these emissions.

  3. Do you know if there are any publicly accessible databases that contain rainfall or other weather information for the Williamsburg area? If you had some of that data, you might be able to include them as covariates (variables that are not of direct interest to you but still have an effect on your response variable) in statistical analyses in order to tease apart what is happening between sites.

  4. Hi Robin, great question! The Keck Lab actually has it’s own weather station with data online, down to 10 minute intervals. Once I get more data throughout the year I am going to analyze my water samples and look at correlation between environmental variables. I’ll definitely be including weather variables in that analysis!