United States Department of the Interior
U. S. Geological Survey
810 Bear Tavern Road, Suite 206
West Trenton, New Jersey 08628
March 12, 1996
Dear liaison committee member,
Enclosed are the liaison committee comments and responses on the FY96 workplan from our December 14, 1995 meeting in Edison, NJ. Attendence was good inspite of the weather and we want to thank all you who could make it. Since it was nearly impossible for the LI folks to make the meeting, we held a smaller information meeting on March 8, 1996 to bring these folks up to speed on our plans. Notes from that meeting also are enclosed.
We also have had an initial review of our workplan by the NAWQA leadership in a 3 hour conference call held February 20, 1996. Because the LINJ study is so different than typical NAWQA studies, afollow up two day review of our workplan is scheduled for March 20-21. Some changes in our plan are likely, for instance, the SW fixed site network have already resulted. We will update all of you with a summary of that review.
With all the furloughs and budget concerns so far this year we find ourselves a little behind where we'd like to be, but we are still moving forward. We would appreciate any comments on our plans so far. Also, over the next year if anyone is interested in seeing how either our GW, SW, or biological samples are collected, let us know and we'll line up a site visit for you. Although a date is not set, we would expect our next liaison meeting sometime early next fall. We'll keep you posted.
LIAISON COMMITTEE MEETING SUMMARY-- December 14, 1995
Surface Water/Ecology Component:
Question regarding studies in the Pinelands. Responded that we spoke with the Pinelands Commission, there is data collection currently, and our emphasis is on urban impacts.
Concern that we need to learn more about the benefits of preserving forested lands and a need to present such information to legislators. Are we planning any forested sites? We responded that not in the fixed sites but in the synoptic surveys.
Concern since Mantua Creek flowed to the Delaware was it still in or out of the GW flowpath modeling area. It just outside the official boundary, but because the Kirkwood-Cohansey was our targeted surficial aquifer and the drainage typified urban development for that system, it was one of our intensive fixed sites and was included in the urban land use survey and in the modeled area.
Are we looking at eastern Long Island and are existing impoundments resulting in questionable interpretations? Discussed comparing forested coastal plain in LI and NJ as possible synoptic study sites, but no plan for fixed sites.
Are landfills and leaking tanks contributing to MTBE or is it only an atmospheric component? Question why are we doing VOCs in surface water because not seen unless near known source? Responded that we are seeing VOCs in more recent data (Hackinsack and LI streams) and haven't eliminated atmospheric and surface runoff sources as a possibility. Urban comprehensive will address this. The modeled half-life of MTBE in the Passaic mainstem is from 7 days in low flow to as high as 40 days in high low, so it will stay around once it gets into the system.
How will we account for white sucker migratory behavior and won't we end up integrating contaminants over the migratory range. Discussed the need to be selective in choosing sites, look for impoundments, choose the proper season, take into account geographical variants, and sample other taxa for comparison.
Clarified the definition of our use of the term ecoregion sites (from the NJDEP benthic invertebrate surveys) to mean reference sites used to establish reference conditions based on ecoregion differences.
Questioned if we could look at pesticides in cranberry bogs in the coastal plain. We don't necessarily want to target specific contaminant uses. We'll look at AMNET bio information and whether these areas are impacted, then decide if a synoptic survey site should be located accordingly.
Wondered if the SW, GW, unsaturated, and atmospheric components of the urban comprehensive efforts would all be included in national synthesis components. Yes, actually is driven by VOC synthesis; unsaturated and atmospheric components will be done at up to 5 study units and SW (urban IFS) and GW (urban LUS) at as many SUs as possible.
Question on how we are prioritizing urban land use surveys vs. study unit surveys. Urban land use surveys and flow path surveys had the higher priority. GW study unit surveys would be necessary for hardrock areas and we'll try to do as much as we can this cycle.
Some felt the critical water supply issues were in the Delaware Basin and Pinelands. Discussed the Delaware R diversion to Camden and eastern Gloucester Counties and need for withdrawals in the Kirkwood-Cohansey to meet projected growth not supplied by the Delaware R diversion-- largely the area we plan to model.
Some wondered if we weren't crossing into Delaware Basin NAWQA territory. Discussed the arbitrary drainage boundaries and our need to look at urban impacts on the surficial aquifers of the Kirkwood-Cohansey. Comprehensive urban, water supply, and VOC issues are best tackled in the Camden metro area, where we'll be doing our modeling and flow path/land use surveys.
Regarding GW task force, did we check with well head protection program to see what existing work has been done. Check for studies done in Camden area and in coastal plain? Yes, Tony Navoy's efforts are the best example and the LINJ efforts will be the next step beyond that. GW flow and contaminant transport modeling addresses aquifer vulnerability on a regional scale. Discussed traditional assessment vs methods development. We will use more sophisticated methods of analysis than previously attempted, including the evaluation of chemical reactions to retard or not retard chemical transport. The modeling will also elucidate where we should target more flow path studies at different depths. Discussed how in southern New Jersey it is easier to demonstrate new tools than working in northern New Jersey? Still, concern for not emphasizing work earlier in the project in northern NJ was voiced.
Question on size of areas to be modeled: 300m2 for LI vs. 400-500 m2 in NJ.
Will we look at migration into PRM? No. Herb agrees it is a lower priority, human signature takes longer to be seen in PRM.
Questioned if rainfall would be sampled. Yes, for pesticides not VOCs. VOCs will be sampled in air samples not rain.
How was Leon's flowpath figure generated, what were the data points, and what was the size of the nodes? Model uses topography, precipitation (recharge) rate, elevation of strata, water use (needs to be updated and has some data missing). Output from MODFLOW feeds into MODPATH to get path and travel time of water.
Will we be using new or existing wells? 1) Urban land use study will use new wells. 2) Flow path studies will use both available wells or new as necessary and appropriate. 3) Study unit surveys will use existing wells or existing data in some cases (aquifers).
Will we be doing flow path surveys/modeling in northern NJ? It is quite an undertaking for flow path study in hardrock or valley fill since will don't have the details of realistic hydrogeology/water use/hydrology to determine flow paths. Coop program may be better way to address this detail.
Need to give technical assistance to decision makers (i.e. Hanover cluster VOC and TCE problem). We need to address how to take data and put it in an understandable format for decision makers who don't understand either the intricacies or technicalities. S. Jersey always gets studied first, for example, well-head delineation in S. Jersey. Worried about availability of FY 98 funding for the N. Jersey surveys if we don't move them up in priority. Team will consider working in a subunit survey in valley fill early next year.
E. Hanover has multiple GW problems, but VOC and TCE not seen in streams. Are we tackling a "too easy" system in south Jersey. Why not try a more complicated system and work in north Jersey this fiscal year. Responded that we were starting with a less complicated system to demonstrate that our modeling works. Passaic model is on the list of cooperative modeling studies, no year decided yet. However, with current approaches Passaic model won't give the answers we need. Non-homogenous flow modeling is needed there and may not be a technical reality to do now. We cannot use general estimates of steady state modeling analysis to obtain a detailed answer in this complex system. Can't link models without additional testing and additional wells. Have to modify and build on existing models.
Where GW is the primary water used in parts of northern NJ, there are major water quality problems. Thinks studying only south Jersey is a problem and that politically we should do something in north Jersey.
Does state have data on VOCs in atmosphere? Total hydrocarbon data only in historical data. Has been talking with Division of Science and Research about VOC sampling.
Is there auto emission data on MTBE? None available. EPA Toxic Substances Inventory doesn't include autos. But air quality is thought to be significant, for example, the study in Alaska found MTBE in blood of commuters.
FOLLOW UP MEETING ON 3/8/96 WITH THE LI CONTINGENT OF THE LIAISON COMMITTEE.
The follow up meeting is the result of the inclement weather that made it impossible for LI folks to attend the December meeting. As it turned out, several LI folks couldn't make the second meeting because of snowy and icy conditions. For those of you who couldn't make it and anyone else, please comment as necessary on both meetings.
As part of the retrospective analysis for the LINJ study unit, Steve Terracciano has compiled data on VOCs in the SW and GW of Suffolk County, LI. This data has been made available by the Suffolk County Health Department and includes analyses for the much talked about fuel additive MTBE as well as 59 other VOCs. Suffolk County, which encompasses the eastern half of LI, is highly developed and urbanized in the west. Development diminishes towards the eastern half of the county where land use is predominantly minimally developed or agricultural. Data analysis efforts are not yet complete. Paul Stackelberg presented preliminary results of early data analysis for SW data on VOCs and pesticides in 93 streams and 582 monitoring wells. The detection level for all VOC analyses was 0.5 ug/l.
Detection frequencies for finding any one of the 60 VOCs in 93 Suffolk County streams during the 3 years was 46%. The compound 1,1,1-trichloroethane (TCA) had the highest detection frequency of 32% and a median concentration of those samples with detections of about 1 ug/l. MTBE detection frequency in the 93 streams was about 29%, the second most frequent VOC detected, with a median concentration of those samples with detections of about 1 ug/l. Still, 56 streams have had no hits of any VOC. None of the 39 pesticide compounds were detected in SW.
The detection frequency for finding any one of the 60 VOCs in the most recent sample of 582 Suffolk County Upper Glacial aquifer monitoring wells was about 25%. Again, TCA had the highest detection frequency of 13% and a median concentration of those samples with detections was about 3 ug/l. MTBE detection frequency in the 582 wells was about 9%, the third most frequent VOC detected. The median MTBE concentration of those samples with detections was about 1 ug/l. Aldicarb sulfone and sulfoxide were the most frequently detected (about 7%) pesticides in GW.
In general, detection frequencies of MTBE and other VOCs in SW and GW appear to be higher in the more developed western part of Suffolk County than in the less developed eastern part. This indicates the pervasiveness of MTBE in highly developed areas of the county. Co-occurrence of MTBE and other VOCs was common.
Information on possible sources of VOCs is also being compiled for each sampling site. Preliminary information indicates that there are no known sources for these compounds for over a quarter of the SW sites at which VOCs were detected during this 3 year period. Since most of the samples probably occur during some stage of baseflow, VOCs in SW could be derived from atmospheric deposition (MTBE) and/or dispersed inputs of GW. There are several sites which have operating GW pump and treat remediation permits located in the drainage area.. The effluent from these pump and treat operations are allowed to discharge into storm sewers at MTBE levels of up to 50 ug/l, clearly suggesting a role of point sources in some of the systems as well.
Currently, we intend to produce a factsheet summarizing this information on VOC concentrations in SW and GW systems of LI's asap. This preliminary analysis of the data is strong evidence, however, for the need to sample all compartments of the hydrologic cycle as proposed in the SW, GW, and urban comprehensive study components of our workplan. The existing data clearly indicates the very high probability of detecting VOCs in all compartments of the hydrologic cycle and the need to address source, transport, and fate issues. The LI folks concur.
These results were followed by presentations/discussions on our urban comprehensive approach which in addition to the SW and GW sampling now includes atmospheric and unsaturated zone sampling as well. Short presentations of what we'll be doing in the Glassboro area in FY96 were followed by discussions of preferred work efforts in LI. A quick summary is as follows.
Atmospheric source sampling of VOCs and pesticides includes 3 sites in FY96-97; one at the Camden NJDEP site upwind of the study area and the other two within the study area, one in a high density traffic area and one farther removed from traffic. VOCs will be sampled in air and to some extent in precipitation; pesticides only in precipitation. When asked who on LI collects the EPA regulatory air quality data, we were referred to Bill Spitz at the NY DEC.
The co-occurrence of TCA is very interesting with respect to the atmospheric deposition component. The atmospheric work has been more of a focus on MTBE expectations, therefore, the more VOCs that show up frequently from atmospheric sources, the higher the quality of our recommendations for unsaturated zone transport because of the selective removal due to microbes. The LI data base is incredibly valuable and suggests that after we get our air quality plans together for NJ, we should consider simultaneous air quality data collection on LI.
Unsaturated zone sampling and modeling includes at least the 3 sites in FY96-97, preferrably co-located at the two atmospheric sites within the study area. VOCs will be sampled in air via soil gas tubes and perhaps pesticides will be sampled on sediment (soil samples from a split spoon sampler). The premise here is that atmospheric loads of VOCs and pesticides will pass through the unsaturated zone in recharge water either conservatively (as MTBE might behave) or non-conservatively (as BTEX would). Our unsaturated sediment sampling plans depend on the outcome of some exploratory work being conducted by Ron Baker and Jessica Hopple in the lab. We need to develop a method of extracting the chemicals of interest (MTBE, TCA, BTEX, pesticides?) to get a detection limit commensurate with total concentrations anticipated for the atmospheric deposition route. Jessica is looking into existing schedules to see if they require modification.
Let's say we find VOCs in the unsaturated zone, then knowing the profile in the unsaturated zone and the atmospheric loading, Art plans to model the movement of detected VOCs through the unsaturated zone to determine the load to the water table. We plan to follow up with a very similar approach to sampling atmospheric and unsaturated zone VOCs and pesticides(?) in the vicinity of the LI flowpath model area in FY97-98.
There are two general types of SW synoptic surveys in NAWQA. One type has the goal to define the spatial variability within SU strata or to compare and contrast concentrations for multiple stream systems. These would be multiple site surveys of different stream systems during predefined flow conditions. The other type has the goal to define sources and mechanisms within a given basin. These would be multiple site surveys of the same stream system during predefined flow conditions. Both types will be used in the LINJ study.
We originally planned but probably will delay until next year a couple of SW synoptics to define the spatial variability of VOCs and pesticides within the major SW strata. Later surveys of the same sites would be directed at sampling biology and perhaps tissues. The workplan currently defines a conceptual synoptic network of 18-23 sites for this survey. All sites currently have nutrient data provided thru the NJ and LI networks. By sampling VOCs, pesticides, tissues, biology, and nutrients, these surveys would help define the relative importance of each on stream biology. These multiple lines of data would be collected at a minimum of 3 streams for each land use. It was pretty clear from Paul's presentation that the VOC and pesticide data base for LI already provides much of this information for LI streams, NAWQA needs only to follow up on biology and tissue data.
Therefore, it was discussed in the meeting that a couple SW synoptics to define VOC sources and mechanisms in 1-3 LI streams would yield some useful information. Three streams having different levels of VOCs in the LI data base would be selected, with at least two of the streams in the flow path model area along the south shore. By sampling these streams below storm sewer outfalls and at other strategic sites during 2-3 flow conditions, we could better define the role of GW and atmospheric sources of VOCs.
We will use the extensive GW data base that exists for LI to help define the loadings from various land uses. An equivalent of at least 5 land use surveys has already been accomplished thru the Toxics studies on LI , including regression equations for predicting detection and loadings. The LI folks clearly indicated that we shouldn't have to carry out an occurrence and distribution study of LI. We agree.
Rather, we will begin to add to the LI understanding later this year through various modeling analyses to evaluate the fate of shallow (contaminated) GW systems in association with identified VOC occurrence in GW and SW and the potential impacts on the deeper water supply source (Magothy aquifer) in the vicinity of the 2-3 stream basins. The approach will be similar to the one currently being applied to the Glassboro, NJ comprehensive urban and flow path modeling study. A 3D GW flow model (MODFLOW) and particle tracker (MODPATH) have already been developed for Glassboro to help define data collection needs and contributing areas to wells. Later, a non-conservative regional transport model will be used to assess the distribution and fate of pesticides and VOCs in the urbanized portion of the Kirkwood-Cohansey aquifer system.
Similarly, for LI a more detailed (finer descretization and layering) 3D GW flow model and particle tracker will be developed for a subarea of the LI regional flow model in the vicinity of the 2-3 stream basins. From this model and a non-conservative transport component, we can conceptualize the flow system (essentially generalizing the deeper Magothy and focusing on the shallow GW/SW system) and thereby evaluate the fate of the shallow contamination on streams and the deeper system. The model analysis will be coupled to the SW sampling and likely to some additional shallow GW sampling in the area to determine role of GW in contributing to VOCs in SW. Coupled with data on the extent of shallow contamination and GW pumpage in the Magothy, generalized temporal scenarios of impacts on the deeper system can be carried out.