Environmental monitoring methods are expensive, labor-intensive, time consuming and often inaccurate. A typical groundwater sampling event requires sample collection, sample analysis, data processing and report preparation to allow for appropriate decision-making and response. This sequence often requires three to six months to generate a final report for each sampling event. This wasteful process is exacerbated by the fact that organic contaminant sample integrity is often compromised during collection, transport and storage.

Response capabilities are severely time limited. For instance, contaminant plume migration to adjacent properties and potential receptors is often not discovered with enough time to respond, which poses significant legal and financial risks. Current monitoring approaches are not always capable of providing real-time analyses of site conditions to enable critical decision-making and contingency response. Furthermore, there is a wasteful carbon footprint associated with conventional approaches. Improved monitoring and conceptualization approaches are urgently needed to allow faster, less expensive and more effective site assessment and restoration, and to enable policy makers and resource managers ample time to make appropriate decisions required for optimized resource management.

Field analysis

Over the past several decades, in an effort to expedite site characterization and remediation efforts, practitioners have developed field analytical methods such as soil-gas analyses, field-gas chromatography testing, ion-trap mass spectrometry, cone penetrometers, and field-test kits, among other so-called "field-screening" techniques. These once-considered innovative approaches are now commonplace and even standardized, as the cost and time benefits have conclusively been documented and even promoted through technology transfer vehicles such as Triad workshops, Interstate Technology and Regulatory Council (ITRC) technical regulatory guides, American Standard of Testing and Materials standards, and federal and state guidance.

Recent advances in integrated circuits, micromachining, micro electromechanical systems, nanotechnology, bio-mimicry, molecular engineering, materials engineering, and computer and electrical engineering have brought forth new opportunities for fielding automated devices capable of measuring chemical and physical properties and characteristics. Robust automated sensors are now available for water level monitoring, chemical monitoring, remedial performance monitoring and even for intelligent data collection control. These new sensors represent the next phase in a logical sequence of industry steps towards expedited measurement and monitoring. While sensors for some of the key organic contaminants of concern are not yet commercially available, several have been in the field for several years.^[1,2,3] Technologies now pervading the groundwater industry – such as the current efforts to develop national security alert systems based on rapid detection of toxic agents through analyte-specific sensors^[5] – will undoubtedly have a direct impact on the future of environmental monitoring.

While robust field sensors will be critical to the future of environmental monitoring, data collection is only one component of an expedited approach that allows rapid decision-making. For instance, sensor data must be reliable and of known quality, must be placed in the proper context for clear understanding, must be organized to produce a graphical representation and conceptualizations, and must be rapidly transferred or made available to the responsible individuals tasked to make and implement critical decisions or prepare essential reports.

The authors have developed and patented (U.S. Patent Number 6,915,211) an innovative process for integrating environmental sensors, telemetry, geographical information systems, and geostatistical algorithms for automatically generating contour maps and models of sensor attributes and multivariate analyses. Groundwater applications include long-term contaminant remediation performance monitoring via flux,^[4] solute plume containment monitoring, vadose zone imbibition monitoring, vapor intrusion, and water resources management and sustainability. A brief summary of selected case studies and future plans for new applications are presented and discussed.

Figure 1: Automatically generated contour map and time series analysis of soil moisture sensor data.

Materials and methods

Figure 2: Groundwater Basin storage change distributions based on dynamic water levels.

Field sensors are linked to data loggers and telemetry systems, which transmit the data to a remote location where it is automatically processed to meet user needs. The system can automatically generate customized reports, and can engage alarms, emergency response plans, and well field and remediation optimization scenarios.

Sensors and sensor platforms are deployed for groundwater monitoring applications in conventional monitoring wells, where they are monitored for multiple constituents and parameters. Many commercially available, multi-sensor platforms can be deployed to simultaneously monitor water level, dissolved oxygen, redox potential, iron species, nitrogen species and contaminant concentration. Sensor platforms generally require water movement, whereas several solid state sensors (e.g. ion selective electrodes) can be deployed in-situ. While most commercially available sensors are connected to telemetry units via cable, others can transmit data to a central datalogger telemetry unit via wireless transmission.

System configuration and operational components are controlled through integration and networking software. Through this package, one can select the type of sensor and telemetry system used, establish display options (examples: background map, symbol and map elements, contour options, time series analyses, color scheme, etc.), control the frequency of data collection, the geostatistical data treatment options, and engage models, alarms, and emergency response protocols through a Web browser. Project management features include document repository, forward projects tracking through geospatial links to Gantt charts, and e-mail tracking. The entire data tracking and reporting system can be accessed through password-protected Web subscription, so no software downloads are required.

Results and discussion

Figure 3: Groundwater seepage velocity distributions based on dynamic water levels.

The examples described below consist of sensor networks deployed at sites under investigation for geotechnical and agriculture optimization purposes, a site where optimized water resource management is a key objective, and at a site undergoing remediation performance monitoring. Customer confidentiality dictates that the authors refrain from presenting contaminant information from specific sites. Since the automated Web-based monitoring system is sensor-neutral these examples serve to demonstrate how contour maps are automatically generated and archived, and how multivariate analyses and models can be applied, regardless of the sensor type.

Figure 1 depicts a contour map of soil moisture content superimposed over a background digital map and sensor locations for spatial reference. This image was automatically generated using data collected on Dec. 1, 2009 at 5:20 p.m. Time series charts ranging from Dec. 1 to Dec. 13, 2009 for selected sensors are also displayed. When practical, virtually any type of sensor that is connected to telemetry can be integrated into the automated contouring package. Multivariate analyses and visualizations also can be automated.

Figure 2 presents a variation on the automated contouring theme, whereby water level sensor data was used to estimate groundwater basin storage change for selected time steps. Water level changes and storage capacity distributions are automatically processed to determine storage change distributions and cumulative volumetric changes. Groundwater divides such as faults are represented to allow simultaneously monitoring multiple basins. This approach enables responsible parties the ability to closely monitor the resource to generate and post reports in a timely manner. Conventional approaches currently require weeks to months to calculate a single incremental basin storage result, while this new approach enables managers to obtain these types of critical reports in a matter of seconds from anywhere with an Internet connection.

Generation of this type of groundwater storage change contour map would require at least several days per map using conventional labor-intensive techniques such as field deployment, water level measurement, data processing, spatial analyses, reporting, etc. The automated approach described enables users to select any two time steps for which sensors have measured data, allowing for analysis of any combination of archived time steps. When analyzing for chemical constituents, generation of report-quality contour maps can often require from three to six months to complete. Alternatively, multi-sensor platforms can be deployed and contour maps for each sensor type can automatically be generated at virtually any step of interest. Furthermore, combined sensor data sets, such as contaminant concentration and redox potential, can be automatically mapped using geospatial analytical capabilities within the graphical information system (GIS).

Figure 3 depicts groundwater seepage velocity distributions based on water level sensor measurements. Previously estimated hydraulic conductivity and effective porosity distributions, which are static, are used to automatically generate velocity distributions every time water level sensor readings are processed by the platform. Similarly, concentration measurements can automatically be converted to mass discharge estimates for automated remediation performance monitoring. Users also can engage a vector feature to display flow direction.

The automated environmental and resources monitoring and management system presented in Figures 1 through 3 is flexible, allowing customized graphical output and default settings. For instance, scales, flow direction arrows, logos, titles and other map elements can be adjusted to meet the project specifications. Furthermore, alternative statistical algorithms can be used, and statistical progeny can be reported and even presented in the graphical images. Data can be exported in jpg, png, shapefile or csv formats.

These examples demonstrate how automated data collection, processing and visualization can greatly simplify and streamline the process of long-term environmental and resource monitoring. Automated sensor and GIS integration benefits include immediate notification of problems such as plume migration to adjacent properties, significantly greater relative understanding of dynamic groundwater systems based on more frequent measurements, sophisticated geospatial analyses such as simultaneous multiple parameter measurement and trend evaluations become cost-effective, less time is required to generate plume and contour maps and reports, greater model precision becomes achievable, carbon footprint reductions can be realized, and decision-quality data can be disseminated to responsible parties with an expedited Web interface. Future plans include demonstration of a nitrate solute remediation system under the EPA ETV program, whereby system performance will automatically be tracked, and collaborators can access the information in an on-demand, password-protected Internet configuration.

Many promising volatile organic contaminant and inorganic sensors are currently commercially available or under development. As automated groundwater monitoring systems and sensors become more commonplace, the sensor and telemetry industry will grow considerably, while more will be learned about how to better protect system components, deploy the systems in stealth configurations, improve upon site specific groundwater models, and ultimately how best to monitor performance of remediation systems.

Automation represents the initial step in the transition toward smart systems, whereby controls, reports, models, decisions, responses, and system adjustments will be automated, ultimately reducing the time and resource consuming steps currently employed in the environmental monitoring and remediation industry. PE

Acknowledgements

The authors would like to express their gratitude to Pablo Bryant and Ram Ray of the SDSU Soil Erosion Research Laboratory, each of the talented collaborators from Trifecta GIS, Dr. Scott Burge of Burge Environmental, Gregg Gustafson of Instrumentation Northwest, Stuart Nagourney of the New Jersey Department of Environmental Protection, Brian Kahl and Cliff Frescura of Groundswell Technologies Inc., and Andrew Barton and Amy Dindal of Battelle Memorial Institute.