Project Title: Inflow Loadings from Ground Water to the Great Bay Estuary
Principal Investigator(s): Larry K. Brannaka, Thomas P. Ballestero, Tom Mack (USGS), and Rob Roseen

Accomplishments
Scheduled Tasks:
During this reporting period, Phases 1, 2, and 3, identification of groundwater discharge zones, flow quantity, and quality characterization, respectively, were to be completed. Additionally, the final work component, building a conceptual groundwater model, was scheduled to begin. Phase 1 has been completed, involving delineation of groundwater discharge zones from analysis of thermal infrared (TIR) imagery. This involved 2 aerial surveys, in April and August of 2000, to evaluate groundwater discharge zones throughout the 50 miles of shoreline encompassing the study area. Over 600 thermal images were inspected for suspected groundwater discharge zones, and then classified and cataloged. Additionally, delineation involved deployment and recovery of instrumentation for calibration data, and image post-processing (mosaicking). Aerial surveys were planned to include one in winter and another in summer. Prior to the surveys, digital thermometers and dataloggers were to be deployed in a calibration array throughout the study area to measure actual field conditions. The collected data would function as calibration data to be applied to the TIR mosaicked images to delineate groundwater discharge zones. Image post-processing would mosaic thermal images into a single continuous image for the study area.

Phase 2 entailed the quantification of groundwater discharge by 2 different methods. A comparison of methodologies for assessing groundwater inflow showed thermal imagery and field techniques are a powerful and affordable tool for evaluation of nutrient loading. This methodology was compared with the more common approach to predict groundwater flow rates using piezometric mapping and aquifer characterization. The first method, involved use of the remote sensing data combined with fieldwork to estimate groundwater flow. The second (validation) method created the most detailed groundwater map to date for the areas immediately adjacent to the great bay and is comprised of nearly 300 monitoring wells. Flow estimates from the two methods were then compared and have shown remarkable agreement.

Phase 3 involved analyses of water quality for groundwater discharge zones and monitoring wells (bedrock groundwater). Based on the water quality results and flow estimates from the two methods, total nitrate loading was estimated for the study area. A cursory analysis of nitrate loading from the Oyster River and the Lamprey River, as well as from the Durham and Newmarket, NH wastewater treatment plants for comparative purposes. These comparisons will help understand the relative importance of nitrate loading from groundwater, surface water, and wastewater.

Progress on Tasks
Phase 1: Delineation of Groundwater Discharge Zones
Phase 1 has been completed with thermal infrared surveys performed in April and August of 2000 by Larry Davis Aviation, along with a UNH researcher. The April survey was performed at 10,000 feet elevation and the August survey at 4,500 feet elevation. The April survey maximized temperature differentials in early spring after the ice had cleared form the Great Bay. Survey conditions were ideal for discharge within the tidal zone: clear skies, low wind, ambient air temperature of 34° F, and an expected groundwater temperature of 50° F. The bay temperature was nearly 45° F, which was less than ideal for locating deeper submarine discharge zones.

Previous research, found in the literature review, has shown that the bulk of the SGWD can be expected within several meters of shore and within the tidal zone. Due to intense tidal flushing it was necessary to perform the aerial survey at low tide. Otherwise, the groundwater thermal signatures would be obscured. Our results have been consistent with past studies reported in the literature, that indicated that most of the discharge zones have been within the tidal zone. The explanation is two-fold: first, the freshwater-saltwater interface of an unconfined aquifer results in a saltwater wedge. The lower density freshwater is pushed upwards along the coastal saltwater wedge. As the groundwater is pushed up and out it creates discharge areas within the tidal zone. The second factor contributing to the occurrence of discharge areas within the tidal zone is the accumulation of marine clays. Clays prevent substantial water from flowing through them, and thereby act as confining units. The perimeters of estuaries are often depositional mudflats. The mudflats typically pinch out at the tidal zone, resulting in less resistance to groundwater flow. The tidal zone therefore represents the path of least resistance for upwelling groundwater. This suggests our survey should have recorded the majority of groundwater discharge zones.

In August, a second, less successful survey was flown, despite many weather related difficulties. The survey flight was forced down to 4,000 ft to obtain acceptable results. Alterations were necessary in the survey plan to accommodate the change in imaging elevation. As a result, image resolution was improved from 15 to 3 square feet per pixel. Weather related difficulties prevent a heavy reliance upon the August survey data. When used in conjunction with the April data, the August data still has great utility. Survey conditions were acceptable, temperature of groundwater, surface water, and mudflats were approximately 48, 69, and 80 degrees Fahrenheit, respectively. Since this was the second survey, specific SGWD zones identified from the previous (April) survey were monitored to determine their respective thermal signatures.

Completion of analyses of the TIR imagery produced a catalogue of suspected groundwater discharge zones. The listing includes a characterization as to size, type, and intensity, and coordinates. For the April survey, a total of 165 groundwater discharge zones were identified along the 150 miles of shoreline surveyed. Table 1 lists some of the groundwater discharge zones. The discharge zones were divided into the following area intervals, based upon field investigations.

Quantification of groundwater discharge by the two methods (TIR and piezometric) proved successful. The two methods use entirely different approaches. As such, the use of the two methods, one new and the other tested and accepted, enable a methodology verification required for widespread adoption by researchers, scientists, and others.

The two methods have shown remarkable agreement. Thermal infrared has the advantage that it can be used to identify exact locations of groundwater discharge, which in some cases behave as point sources. Other flow assessment methods assume uniform diffuse discharge. However areas with a diverse stratigraphy and/or bedrock influence can exhibit a combination of concentrated and diffuse discharge zones. This is true in inland or shallow estuaries where accumulation of marine clays occurs. The accuracy of estimates from piezometric mapping suffers with complex subsurface conditions or limited site characterizations. In these locations thermal imagery can be especially useful as a direct assessment of groundwater discharge, and may provide more reliable estimates. With the thermal imagery, groundwater discharge is evaluated directly, without the need to evaluate or address upgradient factors. Where zones of high nutrient loading are identified, a detailed characterization of upgradient conditions may ensue. These two methods provide a suite of resources with which to characterize groundwater discharge.

Thermal Imagery And Field Techniques
The use of TIR and field methods to assess flow is a direct measurement, previously unattainable without the remote sensing. This in contrast with the more common approach to predict groundwater flow rates using piezometric mapping and aquifer characterization. The remote sensing enables a large-scale evaluation of discharge zones over a short period of time. The entire survey was performed in a few hours. Months of preparation and monitoring of environmental conditions were necessary to ensure optimal survey conditions. The images are useable immediately thereafter with out post-processing. Upon analysis of the TIR imagery, and a subsequent cataloguing of the suspected discharge zones, field investigations were undertaken to assess the reliability of TIR analysis for identifying groundwater discharge zones.

Field verification is performed by locating discharge zones using the thermal images and topographic maps. Field investigations and characterization typically involve verifying the size of the discharge area, confirming an upward groundwater gradient, quantifying the flow per unit area, and sampling the groundwater quality. The discharge water salinity was monitored to verify presence of groundwater rather than saltwater. Typical measured salinity was less than 16 parts per thousand. Using seepage meters to obtain a flow per unit area, combined with area derived from TIR analysis by GIS, a total flow per discharge zone was calculated.

Data Collection and Synthesis
Once the cataloguing of the discharge zones was completed, numerical ranges could be applied based on field verification for a subset of the 165 suspected discharge zones. The discharge zones were divided into the following area intervals, based upon field investigations (see Table 2). This was accomplished from the catalogued data set, which was classified into subcategories of size, type, and intensity; it was necessary to "Eestablish response ranges for each parameter to ensure discrete gradients among responses." This ensures a consistent qualitative data set from which flow measurements can be applied. A method was developed based on one used by the National Estuarine Eutrophication Assessment: Effects Of Nutrient Enrichment In The Nation's Estuaries in which an evaluation of discrete parameters was used to assess water quality over a large-scale. This method used a scoring system, based on numerical ranges, applied to each of the parameters, to integrate a large data set. Based on the classification criteria such as surface area, concentration, and frequency of occurrence, an aggregation could be accomplished to assess large areas.

A subset of suspected groundwater discharge zones were characterized (about 10% of the total), and the flow results applied, in a GIS framework, to the complete set. A similar method as used for the National Estuarine Eutrophication Assessment was applied to determine total discharge to the Great Bay. Table 4 illustrates the following method that was used to determine flow expression determinations. A flow expression determination is a series of if/and/then statements that process data based upon multiple classifications, to obtain individual flow estimates for a discharge zone. While extreme flow ranges are possible within this matrix, from 2,168 to 262,219 gallons per day, not all expressions are observed. The largest flow observed is medium/large (size), medium/high (intensity), at 137,415 gallons per day, and the lowest is small (size), low (intensity), diffuse (type) at 2,186 gallons per day. The flows are calculated with the given ranges in Table 2, and Table 3, then if zone is diffuse a coefficient of 0.3 is applied, and then for all zones a coefficient of 0.4 is applied. The diffuse coefficient is applied to account for the large size and the low volume of flow typically observed with this type of zone. The final 0.4 coefficient is applied to account for the nonuniformity associated with discharge zones. The flow intensity is greatest at the center, where measured, and decreases outward within the zone. Unusually large seepage meters (nearly 20 feet in diameter) were used for flow determination to account for variations within the discharge zone. However the zones are often several thousand square feet. Point measurements of piezometric head were taken throughout a zone to verify flow and support the observation of varying intensity.

The field verification has found TIR to be a reliable method for assessing area of the thermal signature and an excellent method for identification of discharge zones. As is illustrated in Figure 1, TIR image analysis for area determinations compares well with field investigations. Limited field verification was performed due to the difficulty of obtaining proper conditions and the difficulty of assessing large areas. The assessed sites ranged from nearly 2500-4500 square feet. Field assessment of thermal signature area is performed by use of a thermal infrared gun, which measures surface temperature in the same fashion as does the imagery. Similar attempts using thermometers inserted in the soils were less successful. The lack of success was attributed to the need to measure the surface temperature (the top millimeter), which changes drastically with depth. Surface temperatures are affected by shadows, wind, and time exposed by the tides.

With additional field investigations, reliability of TIR for locating groundwater discharge zones has been improved. Specifically, reliability has been improved by discerning features unique to groundwater discharge zones. The August survey was found to be less reliable, due primarily to the less favorable survey conditions. As a result, flow estimates have relied primarily upon the April survey. The review of TIR imagery included over 200 images from the 10,000 ft. April survey and over 400 images from the 4,500 ft. August survey. While survey resolution was dramatically improved with the decreased survey elevation, from April to August, other factors such as shadows, glare and wind interfere with a proficient analysis of the August imagery. The August survey resulted in 425 suspected groundwater discharge zones. The difference in the number of discharge locations identified in the two surveys is based on two major factors: survey resolution, and seasonal variations in groundwater flow. The resolution of the April and August surveys are 15 and 3 square feet per pixel, respectively. As a result, with the improved resolution achieved at lower elevations, a greater number of small and low intensity discharge zones will be identified. A comparison of the data sets from each of the surveys shows nearly a doubling of the small zones, and a subsequent decrease in the other classes. There was a more than double increase in the extra large zones that reflects the improved resolution and subsequent ability to identify diffuse, low intensity discharge zones. Extra large discharge zones (13,765-16,365 square feet) are in all cases, diffuse, low intensity discharge zones without extraordinary flows.

Procedure for GIS Analysis of Thermal Imagery
Delineation of the thermal signature area by GIS is necessary for each catalogued discharge zone, for incorporation with the data collection and synthesis. It is necessary to crop the TIR down to only the suspected discharge zone, and the areas immediately adjacent. Cropping is necessary for the purpose of excluding false positive results, which are the dominant problem interfering with expeditious processing of thermal imaging. False positives are anything that might have a similar thermal signal as the suspected groundwater discharge, such as ponded surface water, tree cover, and deep surface waters. They are predictable and readily apparent to the trained observer, and thus avoidable.

Figure 2 illustrates false positives in thermal imagery. The sample thermal imagery is the product of an aerial survey performed in August 2000 over the Great Bay Estuary, New Hampshire. The survey elevation was at 4,000 feet, resulting in a pixel resolution of 3 square feet. The images are standard polarity: the image spectrum of black to white is hot to cold, respectively. The darkest and warmest objects in the image are the exposed mudflats at low tide, while the lightest and coolest objects are generally the groundwater discharge areas and the tree canopy. This particular survey was flown at noon to minimize shadows and maximize temperature differentials of land surface features.

Figure 3 illustrates the current method of cropping an image to delineate the region of interest within which to run the query. This crop encompasses the mudflats, but excludes false positives such as tree cover and surface water. With the known survey altitude and pixel resolution, the GIS query is used to determine the cross-sectional flow area. The query is performed by converting the image pixels to a temperature-correlated grayscale, and selecting a desired temperature range. This selection will be based on the graphical analysis of a plot of grayscale versus area. The results of the selected discharge area are highlighted in yellow.

With the currently available software tools, the image cannot be cropped (or selected) as a polygon, which would improve the ability to delineate the desired area.

It was necessary to develop a method that is reliably repeatable, from which to base the query selection upon. We have determined that a plot of the pixel data, resulting from the crop, reveals a curve similar to a tangent curve. The lower inflexion point has proven a reliable method for determining the area of the thermal signature. This curve is predictable and will be the key to determining a repeatable area of the thermal signature. The curve has 2 inflexion points with 3 predominant slope components to it: the upper slope, slightly negative, bounded above by a horizontal asymptote; the middle slope, steeply negative; and a bottom slope, slightly negative, bounded below by a horizontal asymptote. Figure 4 illustrates a standard type curve for the thermal image plot and the areas of interest for determining the thermal signature area.

We are only interested in the middle and bottom slopes. This should look like part of a tangent curve tipped on its side. The Lower Slope is determined by first anchoring the line at the highest values of the grayscale corresponding with the lowest area values. This line should have a slightly negative slope and run fairly close to the x-axis. To determine the Middle Slope, the slope was estimated between the two inflexion points. Now the two lines are extended, and where they intersect, a line was drawn perpendicular to the curve. This is the inflexion point correlating with the transition from the thermal signature seepage face, to the plume. Figure 5 illustrates the use of the type curve for estimating the middle and lower slopes, and interpolating the lower inflexion point.

Piezometric Flow Estimates
Flow estimates from piezometric mapping and aquifer characterization have been completed and produced a detailed bedrock groundwater map for the areas immediately adjacent to the Great Bay. Nearly 250 monitoring wells were included, the majority were volunteered for use by private homeowners, and nearly 50 wells were associated with Pease International Tradeport, the site of a decommissioned Air Force base.

There were massive efforts required to survey the locations and elevations of the private monitoring wells and sampling water surface elevations. The total survey effort took nearly 3 years. This involved locating perspective volunteers using tax maps obtained from town offices and maps of water distribution systems from the public works offices. The wells were evaluated, located, surveyed, and sampled for water surface elevations and water quality. The final product produced the most detailed groundwater map developed to date for the area.

Creation of a piezometric map consists of measuring water levels in all participating residents' wells within as short a time period as practical, and using survey data to compute groundwater elevations. The resulting data was plotted and contoured. Aquifer characterization involved the geophysical analyses of suspected target areas, slug testing of monitoring wells, and analysis of pump test data. Interpretation of slug tests and pump tests has provided aquifer parameters for comparison with the results of the geophysics analyses. Directions and estimates of total groundwater flow have been derived from the potentiometric maps of the Great Bay estuary.

Phase 3: Determination of Groundwater Quality
Groundwater quality was assessed in two phases: first by sampling and analysis of groundwater discharge zones, and second by sampling and analysis of monitoring wells. The two phases were for the purpose of ascertaining discharge quality (ultimately for calculating estuarine loading) and to determine potential upgradient source water quality. Monitoring wells were primarily limited to bedrock water solely because drinking water wells are placed at significant depth.

Water quality results indicate the groundwater influx had significantly elevated nitrate levels. The elevated nitrate concentrations were found down gradient of low-density residential areas with little agriculture. The residences are predominantly on private septic systems. Bedrock water quality results also indicate significant elevated nitrogen. Preliminary groundwater discharge quality results have been processed and used to calculate estuarine loading.

Bedrock water quality results need to be analyzed further for statistical reliability. The final sampling event for the monitoring wells was performed in June 2001, and results are only recently available. Preliminary results suggest some potentially alarming levels of nitrate contamination, which is surprising for bedrock wells.

At this point, we cannot link the groundwater discharge and nitrogen contamination with source waters, either bedrock waters or waters derived from unconfined aquifers. Detailed characterization of the flow properties and contributing areas, determination of groundwater residence time, and isotope sampling will be necessary to conclusively link groundwater discharge with source waters. The EPA has identified nonpoint sources such as underground storage tanks, septic systems, landfills, and agriculture practices as the primary sources of groundwater contamination.

Anticipated Success in Meeting Project Objectives in Scheduled Project Period
None of the delays will affect the completion of scheduled project objectives.

Difficulties Encountered
Phase 1: Delineation of Groundwater Discharge Zones

The mosaic efforts were only marginally successful. While this idea is still of vast potential, there are many additional difficulties to overcome. The limited success was for a variety of important reasons: the inordinately large number of images attempted to mosaic, the limited experience of the UNH investigators in image post-processing, the interference of false-positives, and the varying grayscale intensity of TIR imagery. While it was known that a mosaic of so many images would be ambitious, initially it was proposed that the co-investigators (USGS) would undertake the process, of which they have substantially more experience. It turned out, the USGS lacked the necessary GIS expertise and the post-processing software to accomplish this task. University researchers shouldered the responsibility of completing the post-processing efforts including acquiring facilities, labor, and hardware. The Complex Systems Research Center at the University of New Hampshire graciously provided resources and facilities by which to post-process. Despite the efforts, the mosaic became too time consuming and progress too slow to justify continuing, and efforts were stopped. A significant number of aerial images have been pieced together into maps.

However, it became apparent, that regardless of the mosaic success, one the primary reasons for mosaicking was not attainable. A mosaic-wide GIS query for groundwater discharge zones could not be accomplished. Two primary complications prevent successfully querying the mosaic: 1) the occurrence of false-positives, 2) the variation in grayscale from one image to another. False positive are the dominant problem interfering with expeditious processing of thermal imaging. False positives are anything that might have a similar thermal signal as suspected groundwater discharge, such as ponded surface water, tree cover, and deep surface waters. They are predictable and readily apparent to the trained observer, and thus avoidable. Currently, false positives prevent automation of large-scale queries. This complication has given rise to efforts to develop a GIS-based software extension to enable queries devoid of false positives. This parallel project should be completed by August 2002.

The second complication was from a variation in the grayscale from image to image. This is a characteristic of the imaging device. The imager has a self-adjusting contrast that is designed to account for variations in surface features. Unfortunately this varying grayscale intensity results in an inconsistency within the images by which to process a query. This is a common problem with various types of remote sensing. For future imagery, discussions with software developers are ensuing regarding the development of a grayscale correction algorithm that would produce images with a uniform grayscale by which to query.

Phase 2: Quantification of Groundwater Flow

Thermal Imagery And Field Techniques
The primary difficulties to obtaining accurate flow measurements via TIR and field techniques have been: 1) the overly complicated and cumbersome GIS-based analysis of thermal signature area, and 2) the nonuniformity within the discharge within the discharge zones.

The cumbersome process, while slow, has not prevented completion. As a side note, as a result of this need, software is being developed to streamline the process.

The nonuniformity of the discharge zones presents the difficulty to determining flow in the field. The discharge zones are often several thousand square feet and therefore unrealistic to assess flow throughout its entirety. To combat the variations, unusually large seepage meters (nearly 20 feet in diameter) were used for flow determination. Point measurements of piezometric head were taken throughout a zone to verify discharge flow and support the observation of varying intensity. These large-scale operations are perfectly suited for GIS application, thus the challenge being the verification and calibration of the data synthesis.

Phase 3:Determination of Groundwater Quality
No significant difficulties have been encountered.

Preliminary Results
Phase 1: Delineation of Groundwater Discharge Zones
The cataloguing of the August survey has been completed. Comparisons of the two surveys reveal distinct differences, likely due to seasonality, and survey resolution. The differences attributable to seasonality are evident when comparing the same discharge zone at two different times of the year. The previous progress report discusses seasonality in detail. The differences attributable to increased survey resolution were discussed in this report in Section B, Progress on Tasks.

Phase 2: Quantification Groundwater Flow

Thermal Imagery And Field Techniques
Discharge zone flow estimates have been produced using the methods described earlier and modeled after the approach used for the National Estuarine Eutrophication Assessment. These estimates are applying the classification schemes developed from the field verification and characterization of groundwater discharge zones to the entire data set. A sample of the discharge flow estimates is provided in Table 6. This sample list is a fulfillment of the expression determination (Table 4) using the catalogue (Phase 1 has been completed with thermal infrared surveys performed in April and August of 2000 by Larry Davis Aviation, along with a UNH researcher. The April survey was performed at 10,000 feet elevation and the August survey at 4,500 feet elevation. The April survey maximized temperature differentials in early spring after the ice had cleared form the Great Bay. Survey conditions were ideal for discharge within the tidal zone: clear skies, low wind, ambient air temperature of 34° F, and an expected groundwater temperature of 50° F. The bay temperature was nearly 45° F, which was less than ideal for locating deeper submarine discharge zones.

Previous research, found in the literature review, has shown that the bulk of the SGWD can be expected within several meters of shore and within the tidal zone. Due to intense tidal flushing it was necessary to perform the aerial survey at low tide. Otherwise, the groundwater thermal signatures would be obscured. Our results have been consistent with past studies reported in the literature, , that indicated that most of the discharge zones have been within the tidal zone. The explanation is two-fold: first, the freshwater-saltwater interface of an unconfined aquifer results in a saltwater wedge. The lower density freshwater is pushed upwards along the coastal saltwater wedge. As the groundwater is pushed up and out it creates discharge areas within the tidal zone. The second factor contributing to the occurrence of discharge areas within the tidal zone is the accumulation of marine clays. Clays prevent substantial water from flowing through them, and thereby act as confining units. The perimeters of estuaries are often depositional mudflats. The mudflats typically pinch out at the tidal zone, resulting in less resistance to groundwater flow. The tidal zone therefore represents the path of least resistance for upwelling groundwater. This suggests our survey should have recorded the majority of groundwater discharge zones.

In August, a second, less successful survey was flown, despite many weather related difficulties. The survey flight was forced down to 4,000 ft to obtain acceptable results. Alterations were necessary in the survey plan to accommodate the change in imaging elevation. As a result, image resolution was improved from 15 to 3 square feet per pixel. Weather related difficulties prevent a heavy reliance upon the August survey data. When used in conjunction with the April data, the August data still has great utility. Survey conditions were acceptable, temperature of groundwater, surface water, and mudflats were approximately 48, 69, and 80 degrees Fahrenheit, respectively. Since this was the second survey, specific SGWD zones identified from the previous (April) survey were monitored to determine their respective thermal signatures.

Completion of analyses of the TIR imagery produced a catalogue of suspected groundwater discharge zones. The listing includes a characterization as to size, type, and intensity, and coordinates. For the April survey, a total of 165 groundwater discharge zones were identified along the 150 miles of shoreline surveyed. Table 1 lists some of the groundwater discharge zones. The discharge zones were divided into the following area intervals, based upon field investigations.

Piezometric Flow Estimates
Estimates of flow based upon piezometric gradient and aquifer characterization have been completed. Aquifer characterization used several pump tests to determine hydraulic characteristics of the bedrock aquifer, and as such is limited to that component of flow.

To obtain flow estimates from Darcy's Law using piezometric gradients, it is necessary to have regions of uniform gradient. The piezometric surface was analyzed and divided into regions of uniform piezometric gradient (see Figure 7), in which flow was then calculated and summed for the entire study area.

Comparison Of Two Methods
Based upon this listing, preliminary total flow estimates for the bay have been completed for both methods. Table 7 illustrates a comparison of the two flow measurements using the different methods. The TIR flow estimate is still being revised and the number has been falling.

Phase 3:Determination of Groundwater Quality
Groundwater quality results are nearing completion. All laboratory results are in, and need to be reviewed for quality. Preliminary discharge zone water quality has been reviewed to calculate loading estimates for the Great Bay. The loading estimates for groundwater, compared with water quality data from two of the dominant tributaries discharging into the Great Bay, are listed to show the relative importance of each source. Nutrient loading from local wastewater treatment plants is being gathered for comparative purposes. The two surface water sources are sampled at the limit of the freshwater extent (at dams) prior to mixing with the estuary and do not include large wastewater inputs. A review of inputs from surface water, groundwater, wastewater treatment plants, and atmospheric contribution, should cover the major sources and provide useful comparisons.

Bedrock water quality needs to be reviewed and inferences drawn based on presuppositions of the possible source waters (unconfined and confined aquifers) of groundwater discharge.

Tasks and activities for next reporting period

Tasks for the next reporting period
Finally, work will begin on the Phase 3 conceptual model, which will integrate the results of all three phases. Composite maps will be assembled, and incorporated into GIS data layers. Maps and results information will be made available on the Internet, and will be published in journal articles. Project summary results have been and will continue to be presented at scientific and resource manager conferences. Of particular interest is the relevance of the new TIR methodology for use in determining Total Maximum Daily Loads. Project data will be presented in the fall at a conference dedicated to TMDL's. Well data will be shared with the state water well board. A several page project summary will be written and disseminated to project participants, including information gathered for the respective homeowner wells. Data sets will be completed and packaged such that they can be shared and used for other related research.

Work plan to accomplish tasks
The data-gathering phase is complete and nearly all of the data has been reviewed. Continued analysis is planned and conclusions are being drawn. The final step is the writing of final project report.

Concerns or difficulties
Concerns regarding apparent variations in the thermal signature of the groundwater discharge zones suggest that further investigation is necessary to determine the cause. It is suspected to be either the result of seasonal variations in total flow or the result of differences in the survey parameters (resolution, environmental conditions). This is not expected to pose a major problem.

Another concern is the variation of intensity or brightness in the thermal images. The imaging instrumentation automatically and continually adjusts the operating temperature range. It ranges approximately 20° F and adjusts depending on what the dominant feature is. The results were a series of images with varying gray-scale intensities. This poses a problem for uniformly applying a query targeted at specific thermal signal. As there does not yet exist a means for effectively querying a large-scale area and omitting false positives, this is not expected to be a significant problem.

Expenditures
The expenditures to date are within the range of those planned for this stage of the project.

References

Bokuniewicz, H. (1980). "Groundwater Seepage into Great South Bay, New York." Estuarine and Coastal Marine Science 10: 437-444.

Giblin, A.E., A.G. Gaines (1990) "Nitrogen Inputs to a Marine Embayment: the Importance of Groundwater." Biogeochemistry 10:309-328

Johannes, R. E. and C. J. Hearn (1985). "The Effect of Submarine Groundwater Discharge on Nutrient and Salinity Regimes in a Coastal Lagoon off Perth, Western Australia." Estuarine, Coastal and Shelf Science 21: 789-800.

Bricker, S.B., C. Clement, D. Pirhalla, S. Orlando, D. Farrow. (1999). National Estuarine Eutrophication Assessment: Effects Of Nutrient Enrichment In The Nation's Estuaries. Special Projects Office And National Centers For Coastal Ocean Science, National Oceanic And Atmospheric Administration.

Penfold, Erin, T. Loder. 2001.Data provided from study of Surface Water Quality Monitoring for the Great Bay. University of New Hampshire.