WEBVTT 00:00:02.000 --> 00:00:08.000 First speaker this afternoon is Andrea. Yeah, we're on off. There we go. Yeah. 00:00:08.000 --> 00:00:22.000 She works at the USGS as a research hydrologist. And she focuses on, 00:00:22.000 --> 00:00:34.000 Exposure fake transport P fast manages a P fast research lab. That's the Eastern Ecological Science Center and, has the Bachelor's degree from Brown and a PhD from Harvard. 00:00:34.000 --> 00:00:43.000 And she's going to talk with us today about an effort to assess P fast occurrence and background concentrations in hamster soils. 00:00:43.000 --> 00:00:47.000 Thank you. 00:00:47.000 --> 00:00:54.000 Alright, thank you very much. So today I was asked to talk a little bit about our work looking at what we're calling anthropogenic background concentrations in New Hampshire. 00:00:54.000 --> 00:01:04.000 Soils. And so the majority of the talk will focus on that work and then I'll transition and talk a little bit about what the USGS is doing more broadly. 00:01:04.000 --> 00:01:06.000 Looking at groundwater concentrations of PEFASs doing more broadly, looking at groundwater concentrations of P. 00:01:06.000 --> 00:01:29.000 Fass using a groundwater concentrations of PFASs doing more broadly, looking at groundwater concentrations of PFAS using a very simple, very similar sampling design is what we and then finally I'll just conclude with a note about the work that my laboratory does, which is located just an hours drive from here. 00:01:29.000 --> 00:01:40.000 So. 00:01:40.000 --> 00:01:46.000 You need to be stuck. 00:01:46.000 --> 00:02:07.000 Oh, there it goes. Alright. Alright, so of course this work was done in collaboration with, many different people from the USGS, primarily Leos Antangelo, Sidney Ballach, and Joe, and then we did this work in close collaboration with the New Hampshire Department of Environmental Services, Jeffrey Marts, and then we did this work in close collaboration with the New Hampshire Department of Environmental 00:02:07.000 --> 00:02:08.000 Services, Jeffrey Marts, Kate Amos Lesser and Anthony Druin. 00:02:08.000 --> 00:02:14.000 And of course, many other people contributed to this project. So, let's just take a step back and talk about what anthropogenic P fast means. 00:02:14.000 --> 00:02:28.000 So when we talk about background concentrations, background for P fast should really be 0 or close to 0. 00:02:28.000 --> 00:02:42.000 You know, it's primarily a human made compound. And so background should be 0. But realistically, we've all seen that there's P fast and rain, there's P fast in the atmosphere, there's PFAS, it seems everywhere. 00:02:42.000 --> 00:02:48.000 And so I'm terming this anthropogenically fast background for the purpose of this talk. 00:02:48.000 --> 00:03:00.000 And that is what we were trying to understand for the state of New Hampshire. So New Hampshire, was required to set their soil remediation standard rulemaking and initiate that by November first, 2,023. 00:03:00.000 --> 00:03:08.000 You'll note that that was last week, which they did for PFNA, PFOS, and PFHXS. 00:03:08.000 --> 00:03:08.000 And so they came to the USGS asking for support because they needed a lot of information in order to do this. 00:03:08.000 --> 00:03:18.000 And so they came to the USGS asking for support because they needed a lot of information in order to do this rulemaking process. 00:03:18.000 --> 00:03:33.000 They've evaluated or could evaluate several 5 different components and that includes direct contact risk-based soil concentrations, bleaching base soil concentrations, background soil concentrations, which is the main focus of this talk. 00:03:33.000 --> 00:03:52.000 Ceiling concentrations and practical quantification limits. I'm not going to delve into any more details on how they've done this process, but I just wanted to mention that this work was directly translated into, rulemaking by New Hampshire DES, which has been a really exciting process to be part of. So. 00:03:52.000 --> 00:04:08.000 This study is part of a larger project with the state of New Hampshire. The first component was to characterize an propaganda PFAS throughout the entire state of New Hampshire in areas specifically not known to be impacted by local PFAS sources. 00:04:08.000 --> 00:04:15.000 And that is because a lot of work to date has been focused on hotspots, places where we know that PFS has been released. 00:04:15.000 --> 00:04:21.000 But with the declining Mcl's proposed MCLs, and, and statewide regulations, it's become more important to understand what is background in our environment. 00:04:21.000 --> 00:04:38.000 I think it's really important not just for soils like we're talking about here, but for groundwater, for surface waters, for atmospheric contributions because when we go to a site and we get a P fast detection. 00:04:38.000 --> 00:04:44.000 At what point does that mean that there's actually a local source or are we just looking at this anthropogenic background? 00:04:44.000 --> 00:04:55.000 And that's becoming really critical to understand. So the second part of this project was extensive laboratory experiments to understand partitioning. 00:04:55.000 --> 00:05:06.000 We did batch experiments, column experiments. We looked at pH and ionic strength effects. We evaluated soils and biosolids all from the state of New Hampshire. 00:05:06.000 --> 00:05:15.000 We looked at adsorption versus desorption, literally hundreds of samples we evaluated in the lab to better understand partitioning within the state. 00:05:15.000 --> 00:05:26.000 And then we also did a field investigation at 2 different sites. One was a biosolids impacted site, which is on an active farm where they have deposited biosolids for several decades. 00:05:26.000 --> 00:05:35.000 And then the second site was a fire training area, which many of us have come worked on one of those with And so this is just to mention that this is part of a larger study. 00:05:35.000 --> 00:05:47.000 And the part that I'm talking about today, which was of interest to this crowd was mostly the anthropogenic background in the soils. 00:05:47.000 --> 00:06:02.000 So how do we, develop an effective study design? For soil assessment? There are a lot of resources and I think Nikki pointed out some of them this morning that can be used to draw from ITRC as a good example for our state of Michigan. 00:06:02.000 --> 00:06:05.000 There are a lot of resources available now. When we started the study in 2020, there weren't quite so many resources available. 00:06:05.000 --> 00:06:20.000 And so I'll walk through some of the things that we considered when we did the study design and it closely matches what the recommendations are on line that you see these days. 00:06:20.000 --> 00:06:20.000 The most important thing is to consider your study goals of course. How are these data going to be used? 00:06:20.000 --> 00:06:35.000 What questions are you trying to answer that will dictate how your study design is set up? Do those data need to be compared to other data? 00:06:35.000 --> 00:06:43.000 In our case, the state of Vermont and Maine both also conducted soil studies and so we wanted to be able to compare our results to theirs. 00:06:43.000 --> 00:06:50.000 Although our design was was quite a bit different. The sample network design, what scale are we looking at? 00:06:50.000 --> 00:06:56.000 In this case, statewide. The number of sites for statistical significance, if you'd like to stratify your data in any way, look at different soil types. 00:06:56.000 --> 00:07:07.000 Or different land uses having enough statistical power in order to do that. And then I'll just group site selection and restrictions on location. 00:07:07.000 --> 00:07:16.000 So how do you define what an optimal site is for sampling? And if you're looking at a small area, this might be pretty self-explanatory. 00:07:16.000 --> 00:07:25.000 You want to sample that particular area for a statewide study. There are a lot of considerations. What do we consider to be an acceptable location to look at anthropogenic background? 00:07:25.000 --> 00:07:45.000 And I'll get into that. And then for soil sampling discrete or composite samples are 2 very different types of sampling and both have their advantages and disadvantages to considering whether or not you want a more homogeneous picture or you do want those discreet samples that are not composited. 00:07:45.000 --> 00:07:55.000 Is very important sampling depths how you're going to actually process your sample. And then this was mentioned also this morning, you know, what other type of supporting data is necessary. 00:07:55.000 --> 00:08:06.000 Organic carbon, pH, medals, you name it. Potentially there are ways to leverage other emerging contaminants of concern along with the study that you're executing. 00:08:06.000 --> 00:08:14.000 You know, we oftentimes have concern along with the study that you're executing. We oftentimes have blinders on because we're focused on PFAS and I'm absolutely guilty of this as well, but there's a lot out there. 00:08:14.000 --> 00:08:23.000 And so if we're going to do a statewide study. It's potentially advantageous to leverage resources to better understand other things as well. 00:08:23.000 --> 00:08:37.000 And like was described in excellent detail this morning, lab. And field QA QC and. And reporting limits, which was a big issue for us in 2,020. 00:08:37.000 --> 00:08:47.000 So our study goal is jumping right into our specific study was again to look at anthropogenic P fast concentrations in soil shallow soil across the state of New Hampshire. 00:08:47.000 --> 00:08:57.000 In order to do that, we limited our sampling to lands that were classified as forested, troubling, herbaceous, barren, or wetlands. 00:08:57.000 --> 00:09:12.000 So that means we excluded anything that was developed land. Agricultural lands and waters obviously. We also placed a 500 meter buffer around any parcels that had known or potential be fast contamination. 00:09:12.000 --> 00:09:21.000 And we were lucky because the New Hampshire had excellent documentation of all sorts of sites and facilities where they knew KEEPS was being emitted or likely to be omitted. 00:09:21.000 --> 00:09:36.000 Those are all those orange dots that you see across the state. And so that eliminated things like airports, waste water treatment plants, fire training areas, landfills, places we know that have 00:09:36.000 --> 00:09:51.000 One of the most important aspects of our study design was to employ this method called stratified equal area random sampling and what this does is it minimizes the bias and it provides equal statewide coverage. 00:09:51.000 --> 00:09:58.000 And so the way that it works is we gridded the state of New Hampshire up into 100 equal area grid cells. 00:09:58.000 --> 00:10:06.000 And then within those grid cells, there was 1 point randomly selected from each grid cell from which we took a sample. 00:10:06.000 --> 00:10:10.000 So that minimizes bias. It provides statewide coverage. It's also incredibly inconvenient. 00:10:10.000 --> 00:10:25.000 And so when we did this, you know, we knew this going in. It was going to be a challenge identifying at those random sites who the property owners are, how we're going to get there. 00:10:25.000 --> 00:10:31.000 For those who are not familiar with New Hampshire, the southern portion of the state is relatively populated. 00:10:31.000 --> 00:10:40.000 The northern portion of the state is basically forest. And so those random points were oftentimes, you know, in the middle of that country, impossible to get to. 00:10:40.000 --> 00:10:47.000 And so we of course had to get to somewhere accessible. But this was this was our design. It was not convenient. 00:10:47.000 --> 00:10:54.000 It was not easy, but it does minimize bias. 00:10:54.000 --> 00:10:59.000 So at all 100 of our locations, we did sample from 0 to 6 inches in depth. 00:10:59.000 --> 00:11:23.000 And that was to compare directly to Vermont and main concentrations. Who also conducted a study looking from 0 to 6 inches in depth at 50 of these locations we went further to look from 6 to 12 inches in depth to better understand migration of PFAS through the shallow soil to the subsurface and then at 6 locations we did a profile down to 36 inches. 00:11:23.000 --> 00:11:31.000 We analyzed for 36 different P fast compounds. Tapa, the total oxidizable precursor assay, which was nicely talked about earlier. 00:11:31.000 --> 00:11:43.000 We did that at 50 of our locations all from 0 to 6 inches in depth. And then we measured pH, total organic carbon protein. 00:11:43.000 --> 00:11:49.000 Given that there is recent research looking at protein as a, as a sorption mechanism essentially for PFAS. 00:11:49.000 --> 00:11:57.000 Percent moisture and then we did a visual classification of the soils. 00:11:57.000 --> 00:12:08.000 So we cleared our land surface of leafletter, sticks, and so forth. We use D fast free sampling equipment, primarily stainless steel, everything, stainless steel trials, bowls, auguras. 00:12:08.000 --> 00:12:20.000 And then, samples. Up the, at the target location. We're collected from 3 nearby locations and then composited and homogeneized within Restala Stale Bowl. 00:12:20.000 --> 00:12:41.000 We cleaned off our equipment between samples without using any methanol and that actually was incredibly effective. We were able to brush off loose soil, rinse with DI water, scrub with Liquinox with the eye water and then did a very thorough DI water rinse and then finally a P fast-free LCMS grade water branch that was also key fast-free. 00:12:41.000 --> 00:12:50.000 And with all of that rinsing and rinsing and more rinsing, we were all of that rinsing and rinsing and more rinsing, we were able to achieve a clean sampling equipment without generating a lot of methanol waste. 00:12:50.000 --> 00:12:56.000 That we would have to transport across the state. 00:12:56.000 --> 00:13:02.000 Okay, so we did collect 22 equipment blanks. We actually collected a lot more than that. We measure 22 equipment blanks and those were measured for PFAS, TAPA and TOC. 00:13:02.000 --> 00:13:12.000 The way that we took blanks for soil was actually a water blank in which we put the we Essentially, put the water over all of our sampling. 00:13:12.000 --> 00:13:27.000 Equipment measured the water volume so we knew exactly how much mass we were dealing with and send that off for aqueous analysis then we could convert that back to a massive PFAS in case of any detections. 00:13:27.000 --> 00:13:36.000 We did source solution blanks and then several sets of replicates and, 20 sets of matrix spike matrix like duplicates. 00:13:36.000 --> 00:13:39.000 To ensure quality control. 00:13:39.000 --> 00:13:45.000 So, Equipment Blank concentrations were really minimal. And if any, if they were ever detected, they were determined to be unlikely to impact any sample results. 00:13:45.000 --> 00:14:13.000 We had bigger questions about method blank detections that came back from the contracts lab. And, we decided to censor the data if it was less than 5 times the method blind detection and that resulted primarily in censoring of PFBS, which seems to be a problem for the laboratory and some minor instances of one PFHXA and 5 6 to FDS samples. 00:14:13.000 --> 00:14:43.000 The duplicate RPD was, surprisingly good considering how a non homogeneous, or, it's less than 25% for all compounds except for all compounds except for all compounds except for PFTRDA which was less than 25% for all compounds except for PFTRDA which was 27%. 00:15:15.000 --> 00:15:25.000 30% are within those 2 dashashed blue lines. Generally, we see really excellent recoveries across the board. 00:15:25.000 --> 00:15:30.000 I've highlighted a few. Where we see some deviations. These are important to keep in mind and keep track of that because they can provide negative or positive bias. 00:15:30.000 --> 00:15:37.000 In your sample interpretation, Luckily those compounds didn't seem to be particularly important to our study. 00:15:37.000 --> 00:15:56.000 But you know these it's really important to look at the QC data and look at it really closely because if you think you have a non detect but all of your recoveries are 50% you know, it leads to a negative bias. 00:15:56.000 --> 00:16:00.000 So it's important to keep this in mind. 00:16:00.000 --> 00:16:07.000 Okay, so getting to some of the results. What we're looking at here is again compounds on the X-axis arranged in the same way. 00:16:07.000 --> 00:16:26.000 And then, the number of samples all the way up to 100 samples from across the state. Samples that were less than MDL are shown with the gray bars and then in orange are detections with a J flag so that means that they were in between the method detection limit and the recording limit. 00:16:26.000 --> 00:16:31.000 And then detections with no J-flag. In other words, they're above the recording limit or in blue. 00:16:31.000 --> 00:16:40.000 You can see right off the bat that primarily we're seeing a large range of carboxylate compounds and PFOS. 00:16:40.000 --> 00:16:48.000 So, looking at this another way, if we pause the detection frequency on the x-axis here, we have the number of per fluorocarbons. 00:16:48.000 --> 00:16:58.000 And so that means that PEFOSA is 7 and PFOS is 8. And we have detection frequency on the y-axis and we see an orange are the carboxylic acids. 00:16:58.000 --> 00:17:15.000 The closed symbols are 0 to 6 inches and the open symbols are 6 to 12 inch depth samples in blue we have the same for the And so we see for the Ker Box slits that we have a wide range of carboxylic compounds with eye detection frequencies across the state. 00:17:15.000 --> 00:17:43.000 And keep in mind these samples are from Natural quote unquote land use as much as we could. And oftentimes in very forested remote areas, when we look at cellphones, we see that there is this zigzag pattern and that we also see in our groundwater data the USGS has from eastern United States and actually beyond that as well we see the zigzag pattern where the even 00:17:43.000 --> 00:17:53.000 chain length compounds have much much higher detection frequencies than the odd chain length compounds and that likely reflects manufacturing history. 00:17:53.000 --> 00:18:02.000 When we look at detection frequency of other compounds, they're generally much lower than the the PSAA and so I've put those in a table here on the right. 00:18:02.000 --> 00:18:07.000 Which you can review if you're interested. 00:18:07.000 --> 00:18:16.000 So looking at the concentrations of PFAS, we're looking here. The purple again is the carboxylates, the blue or the soft names at the end. 00:18:16.000 --> 00:18:24.000 And then that green bar that you see is Gen X HF PO DA. I just want to point out that the concentrations on this plot are on a log scale. 00:18:24.000 --> 00:18:35.000 And so those are incredibly low concentrations of Gen X. I think it was oftentimes a judgment call whether or not those were real detect or not. 00:18:35.000 --> 00:18:44.000 So treat that data with caution. The other compounds we see of course that PEPOS and PEFOSA are highest in concentration. 00:18:44.000 --> 00:18:52.000 For PFOS it was about 0 point 9 6 nanograms per gram meeting concentration across the state of New Hampshire. 00:18:52.000 --> 00:19:02.000 I've put the common laboratory reporting limit as one nanograms per leader and those of you who are a student might point out that that is no longer true from 1,633. 00:19:02.000 --> 00:19:12.000 It was true at the time we started this 20 this study in 2,02016 33 did not exist yet and that was the reporting limit we were typically given from various laboratories. 00:19:12.000 --> 00:19:23.000 And so we had to work with the laboratories. They actually were able to reduce their detection limits substantially, which is excellent because as you can see, most of our data would have fallen below it. 00:19:23.000 --> 00:19:37.000 And so that was that was really important for the study to be successful. The median MDL, because the MVL varied by day by batch of samples that we sent. 00:19:37.000 --> 00:19:49.000 And also by, by sample itself, the NDL. Median is plotted as those blue lines just to give you a sense of where the data is falling. 00:19:49.000 --> 00:20:05.000 The most, one important point that I should make is that, across the state of New Hampshire every single 0 to 6 and 6 to 12 inch soil sample that we collected had the technical P fast within it was mentioned earlier that other more global soil studies have also found similar, results. 00:20:05.000 --> 00:20:19.000 So we're looking. Very forested places and yet we still see. Substantial, you fast concentrations. 00:20:19.000 --> 00:20:24.000 Visually, this is what the sum of P fast looks like in the state of New Hampshire. 00:20:24.000 --> 00:20:32.000 There are 2 plots here. The first one on the left is showing nanograms per gram of the sum of 35 different PFAS. 00:20:32.000 --> 00:20:40.000 We did drop one compound PF ODA because of core recoveries. And then on the right we see the sum of 35 PFAS as well in Pico Moles program. 00:20:40.000 --> 00:20:49.000 I only show that to point out that really if we're gonna look at songs we should do it on a molar basis, but we've all been trained for PFFs to think in nanograms per leader and nanograms per gram. 00:20:49.000 --> 00:20:56.000 So I've given in and we're gonna use manograms for grand nanograms per leader for the rest of this talk. 00:20:56.000 --> 00:21:03.000 But the distribution does not change very much as you can see. 00:21:03.000 --> 00:21:12.000 When we split this up and start looking at different compounds and different trends on the left we see PSCAs, the sum of 12 PSCA compounds. 00:21:12.000 --> 00:21:31.000 And on the right, the sum of 8 PFSAs. The scales are the same between the 2 figures and we see that there are much higher concentrations for the sum of PFCAs than there are for PF essays and those PF essays are almost completely dominated by PS OS specifically. 00:21:31.000 --> 00:21:39.000 When we look at PEFOA and PFOS, we start seeing very different trends across the state. 00:21:39.000 --> 00:21:44.000 Pefoa, you can see that there's a higher concentration of higher detections in the southern portion of the state. 00:21:44.000 --> 00:21:48.000 There is a note, first of all, there's higher population density in the southern portion of the state. 00:21:48.000 --> 00:21:54.000 And second of all, there is a known large emitter for atmospheric, key fast in the southern part of the state. 00:21:54.000 --> 00:22:06.000 It's unknown how much they contribute to that PFS. Pfoa, total load there, but we do see a statistically significant increase in PE concentrations in the southern part of the state. 00:22:06.000 --> 00:22:13.000 Key Foss on the other hand is much more broadly distributed. You see high concentrations even in the north. 00:22:13.000 --> 00:22:19.000 And we see very, very strong correlations between PFOS and organic carbon. Hey, in the soils. 00:22:19.000 --> 00:22:49.000 Tapa was somewhat surprising for me. I've done a lot of work on DOD sites and looking at at and fire training areas. 00:23:55.000 --> 00:24:08.000 Vermont is plotted in green and the main study is plotted in purple and we can see that generally New Hampshire had much higher concentrations of P fast across the board. 00:24:08.000 --> 00:24:16.000 And, why this is, I think, is a probably nuanced question, that we don't necessarily have the answers to. 00:24:16.000 --> 00:24:23.000 New Hampshire is more popular populated to start with. Or has higher population densities at least. 00:24:23.000 --> 00:24:32.000 I think there are questions about atmospheric transport and load that we don't know the answers to to really fully say why this is occurring. 00:24:32.000 --> 00:24:43.000 We do see a statistically significant. A decrease in soil concentrations as we move deeper into the soil profile. 00:24:43.000 --> 00:24:51.000 So the 6 to 12 inch samples are generally lowering concentration for all compounds. As we move deeper in the soil profile. 00:24:51.000 --> 00:24:58.000 We can see. This also with those 6 sites where we went down to 36 inches when we when we could get it without hitting the water table. 00:24:58.000 --> 00:25:09.000 And so you can see here that typically P fast concentrations decreased with depth below the land surface. 00:25:09.000 --> 00:25:16.000 Okay, so let's talk about what this means. So As has been pointed out, you know, P fast. 00:25:16.000 --> 00:25:20.000 Is fairly prevalent in our soils as we see in New Hampshire and it really leads to this discussion about what that means. 00:25:20.000 --> 00:25:33.000 So I think we have this massive PFAS. So I think we have this massive PFAS soil reservoir that's sitting in our soil surface and this has been proposed in the literature as well over the last couple of years. 00:25:33.000 --> 00:26:03.000 If we do some back of the envelope calculations and we assume that median concentration of PFOS is point 9 6 nanograms per gram as we determine from our study and a soil dry bulk density of one gram per centimeter cube which is very low and a New Hampshire land area that includes all land use types we can calculate that we have 3,400 So to put that into 00:26:03.000 --> 00:26:15.000 context, that's enough PFOS to contaminate over 7,000 years of domestic water use in New Hampshire at a concentration of 4 nanograms per leader that proposed MCL. 00:26:15.000 --> 00:26:26.000 So that's a totally unrealistic scenario I'd like to point out that's not actually what's happening right because it depends on where the water is coming from and and assumes that you have leeching in exactly 4 nanograms per leader. 00:26:26.000 --> 00:26:32.000 It's not a realistic estimate. It just provides some context of how much 3,400 kg means. 00:26:32.000 --> 00:26:43.000 And that assumes, you know, 79.7 million gallons per day. Used in New Hampshire for domestic uses and that includes drinking and food prep and bathing and washing and so forth. 00:26:43.000 --> 00:26:53.000 So starts to put into context how much PFOS and that's just for PEPOS, how much POSS we're talking about. 00:26:53.000 --> 00:26:59.000 So I'm not going to delve into some of them more interpretive aspects of the study and we are still working on that. 00:26:59.000 --> 00:27:07.000 But we do have statistically significant correlations. Using Spearman Row. Positive correlations using Spearman Roe. 00:27:07.000 --> 00:27:22.000 Positive correlations for protein and. Positive correlations for protein and TOC, positive correlations for protein and TOC, as has been found in many studies, negative correlations for pro team and TOC, as has been found in many studies, negative correlations for pro team and TOC, as has been found in many studies, negative correlations with PH and negative correlation with latitude as we talked about specifically for 00:27:22.000 --> 00:27:25.000 the PSCAs like we showed with PEFFA having a higher concentration in the southern portion of the state. 00:27:25.000 --> 00:27:31.000 So I'm going to quickly summarize this and then move on to some of the other work that the USGS is doing. 00:27:31.000 --> 00:27:44.000 But the main takeaways of this New Hampshire background, anthropogenic background study was the use of an equal area grid approach that minimized sampling bias and also provided statewide coverage. 00:27:44.000 --> 00:27:51.000 It, it was not convenient, but it does provide a more robust set of data. 00:27:51.000 --> 00:28:05.000 Pfas were detected in every single 0 to 6 and 6 to 12 inch sample. We have a reservoir of PFAS that is stuck with in the soils and potentially could slowly leach out over time. 00:28:05.000 --> 00:28:13.000 And, we, we find that concentrations typically decrease with depths in the soil, which indicates retention in the top layers. 00:28:13.000 --> 00:28:23.000 Topper results were surprisingly low. Which I think is good news and we are still working on data analysis. 00:28:23.000 --> 00:28:26.000 So, just to bring this, back into context, that was one part of this larger study with. 00:28:26.000 --> 00:28:35.000 Hampshire. We are also still working on interpreting all of our laboratory experimental data, the column experiments, the batch experiments and so forth and our field studies as well. 00:28:35.000 --> 00:28:52.000 I wanted to point out that we also have a pilot study that is now looking at the we also have a pilot study that is now looking at the bleaching potential into shallow groundwater that is now looking at the bleaching potential into shallow groundwater that is now looking at the bleaching potential into shallow ground water. 00:28:52.000 --> 00:29:00.000 So we know that there's P fast in the soil, that is now looking at the bleaching potential into shallow ground water. 00:29:00.000 --> 00:29:12.000 So we know that there's P fast in the soil, but and we know and we' about ready to publish those results as well. 00:29:12.000 --> 00:29:34.000 Okay, so, switching tracks somewhat, but this is related in the sense that the same equal area design that we use for our New Hampshire study was actually taken from a broader groundwater network that the USGS has and they use the same equal area stratified equilibrium grid approach in order to do a well selection to understand detection frequencies of contaminants and principal aquifers across the United States. 00:29:34.000 --> 00:29:47.000 And so in 2,019, the USGS started adding PFAS sampling to the National Water Quality Network for groundwater. 00:29:47.000 --> 00:29:53.000 And also the California groundwater ambient monitoring assessments program. That's the Gamma program. 00:29:53.000 --> 00:30:04.000 Both of these networks do provide long-term consistent and comparable information on groundwater quality. They've been sampled for many, many things for many, many years besides P fast. 00:30:04.000 --> 00:30:07.000 Just in 2019, we added be fast to the mix. And they provide information on our, the nation's ground water quality and the trends over time. 00:30:07.000 --> 00:30:21.000 So the idea is that these networks are sampled on a ten-year cycle in order to understand changes in quality. 00:30:21.000 --> 00:30:28.000 So in 2019, you can see on the map there, we had samples primarily from the East Coast networks. 00:30:28.000 --> 00:30:44.000 We published a paper in 22, and we are working on a follow-up paper with data all the way from 2019 to 23 which includes all of those purple dots that you see there. 00:30:44.000 --> 00:30:53.000 So the national water quality network. Is composed of 82 networks, each with 20 to 30 wells approximately and the depths are targeting the zone used for drinking water and they also target specific land uses. 00:30:53.000 --> 00:31:10.000 So usually targeting one of those 2 things like urban land use or agricultural land use. They're sampled on a 10 year cycle in order to evaluate the cattle scale trends. 00:31:10.000 --> 00:31:18.000 Likewise, the California Gamma program. Is has a similar network design which allows us to sort of mesh that data together. 00:31:18.000 --> 00:31:30.000 And they're also specifically looking at networks for drinking water supply statewide. And so in, 22, we published a paper where we said, okay, we have all this groundwater data. 00:31:30.000 --> 00:31:43.000 It's collected using our equal area approach. We're not targeting specific sources. We are we are just looking at the landscape and looking at what's in our ground water. And so we decided to build a model. 00:31:43.000 --> 00:31:52.000 We ended up using a business regression tree model to better understand what variables are driving whether or not we see P fast occurrence in groundwater. 00:31:52.000 --> 00:32:16.000 And so we compared 57 different potential predictor variables and you can see a long list of them on the screen here, but generally they included geochemical conditions, hydrologic position and this included Tritium, which is a tracer for groundwater age, well depths, landscape sources, which we were able to leverage the work that the EPA had done with their analytical 00:32:16.000 --> 00:32:26.000 toolbox. Or PFAS toolbox. And then we looked at urban land use, natural land use, agricultural land use and so forth. 00:32:26.000 --> 00:32:49.000 And so our model, found that the most important predictors of P fast occurrence in groundwater was the number one predictor was tritium which is a tracer for groundwater age and tritium concentrations coincide with chemical weapons testing in 1,953 which is around the same time period that PFAS started being widely used and likely emitted into the environment. 00:32:49.000 --> 00:32:58.000 And so we have a fairly good tracer of whether or not we think we're going to have key fast in our sample is whether or not it's modern groundwater. 00:32:58.000 --> 00:33:07.000 The second most important picture you can see is distance from the near its fire training area and then things like urban land use are further down on the list. 00:33:07.000 --> 00:33:18.000 We can look at partial dependence. So for example, on the x-axis here, we see urban land use and as you'd increase urban land use you have a greater probability of a fee fast detection. 00:33:18.000 --> 00:33:26.000 So as we can see here, this model did a really good job at predicting a P fast occurrence. 00:33:26.000 --> 00:33:42.000 We weren't quite ready to actually use it to predict the areas where we hadn't sampled yet, but with the incoming data that we have since 2,019 we are now revisiting that and we are currently building a model that is using only mapable factors to try and predict EFASa crashes across the Conas. 00:33:42.000 --> 00:33:55.000 And so we're only using mapable factors, things like land use. And different sources of P fast because otherwise we wouldn't be able to predict the unknown locations. 00:33:55.000 --> 00:34:06.000 And so that is what we are currently working on. And it's all based on this equal area red design. 00:34:05.000 --> 00:34:13.000 Run a B fast laboratory. It's only an hour from here if anybody wants to stop by. 00:34:13.000 --> 00:34:21.000 It's at the Eastern Ecological Science Center and Carneas built West Virginia. We have a high-resolution mas spectrometer. 00:34:21.000 --> 00:34:29.000 We do both targeted and non-targeted analysis. With a variety of matrices and we can work with very very small sample volumes. 00:34:29.000 --> 00:34:32.000 We are a research laboratory so we're not regulatory. We are really designed to help with research. 00:34:32.000 --> 00:34:42.000 Deal with those weird samples that you don't know what to do with and how to measure, we'll work with them. 00:34:42.000 --> 00:34:47.000 And there you see my other lab group members, Zach Hopkins, David Wanlow and work in Begs. 00:34:47.000 --> 00:34:54.000 So happy to talk about that if anybody's curious. So hopefully we have a few minutes for questions. 00:34:54.000 --> 00:35:01.000 Thanks. 00:35:01.000 --> 00:35:07.000 Great, anyone in the room? 00:35:07.000 --> 00:35:14.000 Okay, I believe we hope. I see a question in the room. Go ahead. 00:35:14.000 --> 00:35:21.000 There's a question here. You got the mic right there. 00:35:21.000 --> 00:35:41.000 Okay, great presentation. The findings that you presented were pretty stark when you consider like for So the ground water protection criteria are less than your median concentrations for P. 00:35:41.000 --> 00:35:47.000 So it really does seem like your recommendation to better understand. The nature of groundwater impacts is imperative, right? 00:35:47.000 --> 00:35:56.000 We've got pervasive soil impacts above a potential. So the groundwater pathway criterion. 00:35:56.000 --> 00:36:03.000 And, you know, the question of background now is going to become Uber complicated. 00:36:03.000 --> 00:36:03.000 I agree. That's a great impact. I think it really, it's becoming really critical. 00:36:03.000 --> 00:36:16.000 That we better understand what's happening in this anthropogenic background that we better understand what's happening in this anthropogenic background scenario. 00:36:16.000 --> 00:36:34.000 We've done so much incredible work looking at hotspots and you know where we know key pass were released but you know it's and this is You know, we know P faster and polar bears and in the oceans, but these concentrations are they're not they're not in substantial and and I think that it really warrants a lot more research on the topic. 00:36:34.000 --> 00:36:40.000 Absolutely. 00:36:40.000 --> 00:36:46.000 So I just, maybe it's a very small point, but you showed, our bias recoveries for both. 00:36:46.000 --> 00:36:55.000 And I was just gonna ask to do if you had idea of what the cause might be. You know, so. 00:36:55.000 --> 00:37:01.000 I'd probably say we this is probably a question for the contract laboratory. I was following EPA method. 00:37:01.000 --> 00:37:08.000 5 37.1 modified and I know EPA absolutely hits it when I say modified EPA. 00:37:08.000 --> 00:37:12.000 37.1 modified and I know EPA absolutely hits it when I say modified EPA 5 37.1. 00:37:12.000 --> 00:37:16.000 But From an analytical perspective, FAFSA is hard, right? Because FAFSA also has this sort of weird transition. 00:37:16.000 --> 00:37:28.000 There's no qualifier ion in many cases. Qualifier ions were not used for P fast or POSIT quantification and so there could be some bias. 00:37:28.000 --> 00:37:32.000 And we're not used for PEPS or POSTA quantification. And so there could be some bias from that. 00:37:32.000 --> 00:37:35.000 Okay. 00:37:35.000 --> 00:37:40.000 Craig, define arcadis just to follow up to Joe's question on, significance to, you know, potential impacts to background or groundwater. 00:37:40.000 --> 00:37:57.000 What, what do you think this has to do? With our concept of background load and surface water. I'm thinking about overland flow, overland flow, storm water driven impacts, you know, this very diffuse. 00:37:57.000 --> 00:38:08.000 Shallow soil impact. Right. Is, is there a need for background surface water and storm water, attributed to, this source? 00:38:08.000 --> 00:38:23.000 Right. I okay, so I don't know if I'm supposed to repeat the question, but I wasn't doing that, but the question was whether or not we should be looking more at a background or genetic background concentrations of storm water or overland flow and surface waters and I think the answer is yeah we probably should. 00:38:23.000 --> 00:38:31.000 The USGS did start doing some sampling of streams and rivers across the US. 00:38:31.000 --> 00:38:44.000 Just this past summer to better understand what's in our surface waters but You know, I think there's an incredible amount of work to be done on that. 00:38:44.000 --> 00:38:49.000 And I think, you know, there was. Talk about the, groundwater surface water interactions. 00:38:49.000 --> 00:39:05.000 I think that becomes really important because everything from what I've seen from my work is that anything that has sort of what we call modern groundwater or water that's interacted with the atmosphere has the potential to have P fast and and sometimes at significant concentrations. 00:39:05.000 --> 00:39:19.000 And so how that all interacts, I think, is. Wildly not. Well characterized at this point. 00:39:19.000 --> 00:39:24.000 I'm a 00:39:24.000 --> 00:39:37.000 Hmm. Okay. Yeah. 00:39:37.000 --> 00:39:57.000 Of the Hey. Oh. It's a Hello. Yeah, Yep, great, great comments. 00:39:57.000 --> 00:40:06.000 Okay, the first question was. Our assumption here of a soil dry bulk density of one grand percent in meter cube and I've, you know, that's a very low value as was just pointed out and I 100% agree that is a very low value. 00:40:06.000 --> 00:40:15.000 We were doing a conservative estimate. So that is the only I found the lowest value I could across the literature and I use that. 00:40:15.000 --> 00:40:20.000 So that's the only reason we chose one. And then the second, comment was about the 7,000 years and what assumptions went into this calculation. 00:40:20.000 --> 00:40:44.000 And again, I will repeat that this is not a realistic calculation. It's just meant to show if you looked at 3,400 kg of PFOS and soil and you just happened to distribute all of that peakos into water at exactly 4 nanograms per liter, which we can all agree is unrealistic. 00:40:44.000 --> 00:40:49.000 How much water are we talking about being contaminated? Just to put into perspective. So, 100% agree. 00:40:49.000 --> 00:40:56.000 It is not a realistic calculation. It's just meant to put things into context. Yeah. 00:40:56.000 --> 00:41:08.000 We have an online question. The online questioner asks, did you consider air emitters outside of the state where prevailing winds could transport PFAS into the study area. 00:41:08.000 --> 00:41:22.000 Yes. So great question and, the answer is that we would need to do an entire study on atmospheric deposition, which comes to one of my favorite points to talk about these days, which is that we really need that. 00:41:22.000 --> 00:41:30.000 And yes, so we, were looking at the soil, which is, you know, obviously linked to the atmosphere. 00:41:30.000 --> 00:41:38.000 We did not do a full atmospheric study. I think that that would be fantastic and I would love to be part of that. 00:41:38.000 --> 00:41:48.000 And ideally also beyond just New Hampshire but nationwide. Yes. Okay.