Speaker Info

Doug Moyer
Supervisory Hydrologist
United States Geological Survey

Email: dlmoyer@usgs.gov


Douglas L. Moyer is a supervisory hydrologist with the U.S. Geological Survey, with over 15 years experience in nutrient and sediment transport within the Chesapeake Bay watershed. He serves as a Principal Investigator for the determination of nutrient and sediment loads and trends for all monitoring stations in the Chesapeake Bay Nontidal Monitoring Network.

More info...

Seminar Abstract

The U.S. Geological Survey (USGS), as a partner of the Chesapeake Bay Program, is responsible for determining the extent to which nitrogen, phosphorus, and suspended-sediment loads delivered to bay from the monitored-nontidal portions of the bay watershed. This is accomplished by analyzing water-quality observations from the nine River-Input Monitoring (RIM) stations to estimate nitrogen, phosphorus, and suspended-sediment annual loads and trends using Weighted Regressions on Time, Discharge, and Season (WRTDS). The resulting trends in nitrogen, phosphorus, and sediment loads are flow normalized to account for the year-to-year variation in river discharge; thus, the remaining trend is a result of changing sources, delays associated with storage or transport of historical inputs, and/or implemented reduction strategies.

Nitrogen, phosphorus, and suspended-sediment loads are showing measurable improvement at many locations across the bay watershed from 2005 to 2014. Trends in nitrogen loads are improving at 44 of 81 (54 percent) NTN stations analyzed. The median reduction in nitrogen load, for these 44 NTN stations, is 0.68 pounds per acre or 10 percent. Trends in phosphorus loads are improving at 41 of 60 (68 percent) NTN stations analyzed. The median reduction in phosphorus load, for these 41 NTN stations, is 0.11 pounds per acre or 24.7 percent. Trends in suspended-sediment loads are improving at 29 of 59 (49 percent) NTN stations analyzed. The medina reduction in suspended-sediment load, for these 29 NTN stations, is 221 pounds per acre or 29.4 percent.

Seminar Transcript

>> Thank you, Scott. As Scott mentioned, my focus today is to go over what the nitrogen, phosphorous and suspended sediment loads and trends are across the Chesapeake Bay Watershed. We've heard in our discussions today within the STAR workgroup and we know -- based on the storms that came through the watershed, we've got many technicians out collecting water quality samples across the watershed, and what I'm trying to do is how do those observed results translate into the load information that we use for many different aspects across the Chesapeake Bay Watershed, and then ultimately what are the trends within those observed data. So my objective today is to communicate the latest, basically everything through 2004, nitrogen, phosphorous and suspended sediment load and trend results for stations in the Chesapeake Bay Nontidal Network. And I really want to focus on two questions, as Scott was mentioning. We're going to break everything up into first how are nitrogen, phosphorous and suspended sediment loads responding to restoration activities and changing land use across the Chesapeake Bay Watershed. So from this we want to turn out into the watershed and say are we seeing an effect of the different management activities, changes in land use, changes in sources, within our load and trend results. Then we want to turn our focus to the bay itself and the loads that are being delivered to the bay from the nontidal network. And the question that we're answering there is what are the trends in nitrogen, phosphorous and suspended sediment loads being delivered to the bay from nontidal portions of the watershed. Now, I want to pause just for a second and say we really can't answer these questions without the team of individuals that really support the nontidal network and the data collection, the funding of the network through the different agencies that are listed here, and then the computation of the loads and trends. So I just want to acknowledge that it's not just me up here communicating on the work that I do alone. There's many others that really help support that activity. And one team that often gets overlooked really are the field technicians that go out and collect the discharge information, the water quality information that is absolutely germane for us to compute loads and trends. So at each of our stations -- and I'll get into a description of the nontidal network. There's about 117 stations in the nontidal network. Our field technicians collect 12 monthly water quality samples and eight stormflow samples. So across the network that's 2,340 observations that are collected each year. Those are the data that drive our analysis. And I'll just point out every now and then we have a couple pictures of those field activities. This is a team from Virginia that's collecting water quality at the Rappahannock, and then here's some of the equipment that we use to get that representative sample from our rivering system. So now I'm going to turn my focus just for the next couple slides is to our techniques for computing loads and trends. Very basic. Really, we want ultimately to compute a daily load, and to do that we need a daily concentration and a daily discharge. Well, our stream gauges give us or daily information. We're continuously monitoring what streamflow conditions are. But I just mentioned we only collect about 20 samples per year for water quality, so we've got to figure out how do we get to that daily concentration so that we ultimately can get to that daily load. So the method that we're using right now is option one. We take our discrete measurements, and these are used to construct a weighted regressions approach that Bob Hirsch developed based on relations between those concentration data and time of year that they're collected, discharge, and then the season as well. That methodology has been published, and it's Hirsch and others in 2010. And then two reports have come out that basically describes the use of WRTDS within the nontidal network; Moyer and others 2012, and then Chanat and others was just released here in 2015. But ideally what we would do is collect water quality concentrations all the time and get away from our statistical model, so that's why I put on here our ideal options, which is where we do so monitoring within a network evolving. Either we use discrete measurements, such as continuous nitrate or other nutrient constituents, and we continuously monitor to those, or we use a surrogate approach through field parameters. We can build relations with those water quality constituents and then repeat those at a continuous time step. But turning back to the use of WRTDS, this is a powerful tool to take our discrete measurements into what our load and trend results are across the watershed. So I want to give you an example of what we get out of WRTDS. So out of the 20 samples collected per year at for this example, Potomac River at Chain Bridge, and this is for total phosphorous, we generate an annual load, for total phosphorous -- and on the y axis we have load in millions of pounds per year. On the x axis we have the years across. And what we see is each year we have -- each dot represents what the annual load is for that year. And what we see is there's considerable variability as you move from one year to the next, and that's really the confounding effects of streamflow. Some years are wet years, some years are dry years. What I want to point out is that during the drought of 2000 through 2002 in the Potomac we see loads were considerably reduced. However, once you get to 2003, or the winter of 1996, we see increased amounts of total phosphorous coming down or the highest loads that we've seen. So what we've struggled with for years is this variability within streamflow really masks the trends that are present within the water quality results. So what we can do through WRTDS is integrate out the year to year variability and flow, and what's left is what is that underlying trend within water quality as a result of change in land use, change in sources, management activities or natural lags within the system. So once we integrate out, we get what we call a flow normalized load. And then to determine what the trend is within phosphorous, we basically do a point to point slope and say what is the change from 1985 to 2014, or the most recent ten years, what's the change from 2005 to 2014. And what we see for this example is that over the long term, 1985 to 2014, we have about a 25 percent reduction in total phosphorous load. The more recent years, the more recent ten years, we have about an 18 percent reduction. So as I'm communicating the trend results, know that it's being derived from this flow normalized response over that time period. So now I'm going to turn the focus up to what's going on within the nontidal network, answering the question how are nitrogen, phosphorous and suspended sediment loads responding to restoration activities and changing land use. As I mentioned earlier, the nontidal network is comprised of 117 stations, but these stations have come on line over different time periods. Starting in 1985 we have our nine major RIM stations, so those are the outlets of all the major tributaries draining to the bay, and then we had 21 long term stations throughout the watershed. In 2004, based on a PSC agreement to expand the network and make consistency on the data that we're collecting, we added an additional 57 stations. And then in 2011, based on the TMDL and how can we use the nontidal network to better characterize improvements and better data across for the feeding of the TMDL process, we added 30 different stations. So for our analysis we need at least five years of data to compute loads, and then ultimately ten years of data to compute trends. So we're really relying on the 87 stations that are listed on the first two items. Also, we have a range of sizes of basins across the watershed. We see that the smallest ranges from one up to the largest of 27,000 square miles within this watershed. So we're really charactering very different types of basins within the network. If we break those stations out by states, you get the following setup. So for the analysis of loads and trends we have 87 stations, and we see that across Virginia, Pennsylvania, Maryland we have 32, 20 and 23 sites respectively. And then as you get into West Virginia, all the way down through Delaware, you see that we have a reduced number of stations. But for a total of 87, all of these sites have at least load information and then for a majority of the sites we have trend information as well. Based on that TMDL expansion, you see where we're expanding the network. We're adding an additional 30 stations. We see that 11 are coming on line in Pennsylvania, seven additional sites in Maryland, and then you see the rest. So just wanted to give you an idea of by state where these stations are located. So now for the next several slides I want to get into what are the water quality responses for total nitrogen, and I'll go through the responses for nitrogen, then for phosphorous, then for sediment. And what I'm going to communicate with each of these, we're going to share two pieces of information; what are the loads across the watershed, and then what are the trends within these loads that we can relate to changes within the watershed. So for total nitrogen, first thing I want to show is what's the distribution of loads that we're observing across the watershed. And when I talk about the loads, first we're taking an average of 2005 to 2014, and then we also want to compare between the sites so we normalize those loads and divide by the drainage area, and that's why we're presenting per acre loads so that we have comparable comparisons for each station. So what you see here is we've got three categories of load information. The blue boxes represent loadings that are of the lowest category and they range any pounds per acre, less than 6.88. And there's 52 of 81 stations. The yellow boxes represent our medium loading basins, and those are basins that generate anywhere from 6.88 to 13.75 pounds per acre. And then the highest loading basin are represented in the red, and those are basins that generate greater than 13.76 pounds per acre. One thing that your eye's drawn to is where are the highest loading basins, and we see for nitrogen most of that's concentrated in the lower part of the Susquehanna, the central portion of the Potomac, and in the eastern and western shores, and we do have one site down in Virginia. So the natural question is, we have an information as to where the loads are, how much are they changing. So what we can do is bring in our WRTDS trend information and overlay that with where we see the different distributions of loads. So what we just brought in is these directional indicators. So upward pointing triangles indicate that the system is degrading. Down reporting triangles indicate that the system is improving, that the loads are reducing. And then dots represent that we have no detectable trends. So of those categories we see that 44 of the 81 stations are showing improving trends. Things are getting better. We're seeing a response to changes within the watershed. Of the 81, 22 of the stations are showing degrading trends, or about 27 percent of the stations. And then 15 of the stations are showing no trend at all. So there's different ways that you can look at this. What I've broken out here for this slide is of those highest loading basins, what are the patterns within those basins. So we've got 14 of these red stations that are indicating that they generate the largest amount of nitrogen across the watershed, and of those 14 we see six are showing improving directions. They're starting to improve. Three are continuing to degrade, and four have no trends. So now what I want to do is show you a different way of looking at these trends. We want to get away from the directional indicators. There's so much information behind them. So what I've done is I've brought out an example of how we're looking at trends as far as how much mass has been reduced or increased in each of these different basins. So first, this is an example from the Susquehanna Watershed. What we have are all of the stations within the Susquehanna listed in downstream order. So here's Unadilla up in Rockdale, New York, all the way down to Susquehanna River in Conowingo. On the x axis we have the total mass either reduced or increased for nitrogen in pounds per acre. Now, the green bars indicate stations that are reducing nitrogen. The orange bars represent stations that are showing increased nitrogen over this 2005 to 2014 period. And then the gray bars show that there's no trend at these locations. One other thing to point out is the number at the end of each of the bars. This is the percent change. So it gives you an idea of how big of a change is this reduction in mass at a particular site. So, for instance, the Conestoga is showing the largest reduction at about five pounds per acre over this ten year period, but it's a 15 percent reduction. Whereas, Susquehanna at Conklin is only showing about a two pounds per acre reduction, but that's a 35 percent reduction. So take that into account. It's all relative to what the starting loads were for 2005. So what I want to do is add these trend results into the rest of the nontidal network. So what we've done, here's the Susquehanna up here, and then we've added in stations from the eastern shore, the western shore, the Potomac and Virginia. This is a total of 81 stations. And I really want to pause and just say this is a big deal. We've got a TMDL that's going on within the network, and this is probably the first of its kind network, where we have observations at this number of stations where we can come up with a trend in load. So we're managing load and now we have a measure of response within that's based on our water quality observations. So what we see, we know that the majority of the sites based on our previous results are improving, so that's really what this graph is indicating. We see that the range of improvements are anywhere from .1 to 5.07 pounds per acre, and the median reduction is about .68 pounds per acre, or about a 10 percent reduction within all of these improving sites. For the degrading stations we see that the range of improvements are anywhere from .04 to 1.21 pounds per acre, and then that median degradation is about .33 pounds per acre, or 7.84 percent. And for those interested, and we'll get into this a little bit later, you can download this figure from the web page. So now I'll go into our phosphorous results. And again, we're going to keep the same perspective. We'll go over the loads and then what are the trends within those loads. So what I've started with now is that combined map where we have the loads overlaying by the directional change. And again, we see a very somewhat similar pattern to what we saw in nitrogen, but one thing that sticks out is we see that some of the elevated loadings from phosphorous really are along the eastern portion of the Chesapeake Bay Watershed. Some of the highest loading basins, indicated by the red squares, are located in the southern part of the Susquehanna, and then we have some down in Virginia. So basically from this map you could get that the total range in total phosphorous loads are .13 to 2.3 pounds per acre, with an average load of .52 pounds per acre. If we look at what the trends are at those sites, we see that 41 of the 60 stations show improving conditions, and there's 12 of 61 stations that are showing degrading conditions. And out of those additional sites, seven of the 60 are showing no trends. So if we turn our focus to the high yielding or high loading basins, there are six in that category and we find that four are showing improving conditions, and then one is continuing to degrade. Now, if we take those over to our bar plot here, we really see that the vast majority of the stations are showing improving conditions for phosphorous. This is huge information or great news for the Chesapeake Bay Watershed, seeing this reduction in total phosphorous. The largest reduction that we're seeing right now is the Rapidan, which is coming in right around one pound per acre reduction over that time period. If we look at the range within all of these improving stations, we see that we have anywhere from .014 pounds per acre to 1.08 pounds per acre improvement. And then for the degrading conditions the range of degradation is anywhere from .01 to .43 pounds per acre. Now, for sediment, if we look at the loads -- again, sediment follows a very similar pattern as to what we see for total phosphorous. We see more of a scattering pattern for the higher loading basins, but there still is, along the eastern portion of the basin, more of your medium to high loading basins. So the range of sediment loads across the watershed is anywhere from 18 to 2,206 pounds per acre, with an average loading of 482 pounds per acre. For whether these basins are improving or degrading, we see 29 of the 59 stations are showing improvement; whereas 19 of the 59 stations are showing degradation. Turning our focus to the highest loading basins, there are seven in that category. We see that three are showing improvements; whereas, one is showing continued degradation over this time period. And then putting it within this context, we see the distribution of changes across the watershed with the range of improvements anywhere from 8.11 to 1,490 pounds per acre. And then for the degrading sites we see anywhere from 4.75 to 341 pounds per acre on that degradation.

Seminar Discussion

Coming Soon