Lagrangian Sampling? Application to Environmental Studies in an Ever-changing System
One of our poster presentations at the Georgia Water Resources Conference this year was about a research project to find the source of high bacteria counts in one of our local streams. The focus of this presentation was the unique sampling scheme we used to collect water samples called “lagrangian sampling.” Prior to the project, we had collected water samples at different places on the stream to find the place where bacteria counts were highest. We found high counts in multiple places, but no one site consistently had higher counts than all the other sites. We determined the source of bacteria was not confined to one specific location, but instead, appeared to be spread out across a large area. This is known as a “non-point source.”
To better understand where the bacteria were coming from, we applied a lagrangian sampling approach. The basic idea is that you collect water samples from the same “packet” of water as it moves downstream. The way we do that is by adding a special dye to the creek that we can track with a measuring device. When we detect the dye downstream, we know we are seeing the same packet of water and it’s time to collect our sample. The more traditional approach would also involve taking water samples from different places in the creek, but without targeting the same packet of water. Lagrangian sampling takes a lot more time and effort, but we think the data we collect this way is far more valuable. When we take samples from a moving body of water like a stream or river, that water sample has been influenced by the conditions it experienced on its path to our location. If you sample different packets of water, they may have experienced different conditions. By sampling the same packet of water, we can compare a water sample with another from a site upstream, and know that any differences are a result of something happening between the two sample locations.
Experience has shown us that bacterial counts tend be highly variable, even among samples from the same times and places. A lagrangian-style sampling approach for bacteria helped to reduce variability from confounding factors, and in this case, produced a surprising result. We found an extremely strong correlation between the amount of bacteria present and the distance downstream of each site. If there were some entry point of bacteria to the stream, you would expect to see high counts at that point, and lower counts moving further downstream. What we saw was the amount of bacteria increased at a consistent rate as you move downstream. What does that mean about the location of the source and how it’s being transported into the stream? One hypothesis is that the bacteria could be living in the stream bottom and the streambed itself is the source. There are studies in the scientific literature suggesting that bacteria can survive in sediments longer than previously thought. The bacteria in question are not necessarily pathogenic (disease-causing) themselves, but are often used as indicators of potential health hazards due to waste contamination. This would beg the question, “Are these good indicators of contamination and health risk?” The results of this study have given us new perspectives and unique insights, and in this case may have revealed more questions than answers.