Traditional field measurements for water quality and stream gauging still largely depend upon the pencil and paper field notebook for data collection. Although simple, this method is labor-intensive and susceptible to recording and georeferencing errors during transcription. Recent advances in scalable wireless telecommunications and the proliferation of hand-held computing devices has enabled the development of software tools to assist environmental scientists in improving the collection of field data. This paper describes an electronic journal which has been developed for environmental and geolocation data collection. The objective of the mobile software application is to streamline the collection process and improve the accuracy of environmental field data as compared to current practices. The integrated system includes data collection from a water quality probe and a global positioning system sensor, mobile geographical information system software, an internal database and manual input of hydraulic and water quality parameters through a customized user interface. In addition, wireless and Internet technologies are used for transferring and displaying collected field data. Field data mapping using Esri ArcPad, Esri ArcIMS and a customized .NET GIS web service were also developed for enhancing data collection and visualization for the field and remote user. A prototype system has been completed and tested in field trials in Cambridge, Massachusetts, USA and Newcastle, New South Wales, Australia. Field studies demonstrate the noticeable gains in efficiency and precision achieved with the use of the field streaming technology. Potential applications for the electronic field notebook include biogeochemical studies, hydraulic and hydrology studies, watershed management studies and emergency response to water-borne disasters.
The development of software applications for mobile computers has been recently spurred by the availability of more powerful operating systems and the transfer of standardized programming languages on ever-smaller computing platforms. These developments have opened the door for creating new applications that bring computing power to field scientists (Carver et al., 1995; Pundt et al., 2000). One potential application area explored in this paper is the collection of environmental and geopositional data. Associating environmental parameters with a location is essential in many earth sciences. For disciplines with an active field work component ( e.g. water resources, geology, biology), providing software tools that improve the accuracy, efficiency and quality of the data collection process can significantly improve current practices. In addition, the technology in this paper allows users to link real-time field data collection from various devices to a centralized data server located at a remote location, a concept termed field data streaming. This capability can lead to improvements in the management of deployed field teams.
A prototype for the integrated field data collection system has been developed in the Department of Civil and Environmental Engineering at the Massachusetts Institute of Technology. The prototype system, known as the ENVIT Integrated Field Data Collection System, consists of various mobile and wireless technologies that used in conjunction provide a robust, expandable platform for mobile data collection from environmental sensors and manual data input. The ENVIT system is a combination of hardware and software components on various platforms, including mobile computers, a field server and a remote web server. The elements of the ENVIT system include:
The objective of the integrated mobile system is to streamline the collection process and improve the accuracy of environmental field data. From the field site to the office, the collection and automated transfer of environmental data associated with a location can be seamlessly executed without the loss of precision that typifies transfers across various media. By using mobile GIS technology, the system adds a new dimension of spatial localization to the data collection process and provides the user with both textual and spatial cartographic displays. Finally, this integrated system can be used in the coordination of a field study from a centralized location and real-time data display via a web server.
The area of application explored in this study is the collection of water quality, hydrologic and geospatial data as part of a watershed study. To this end, the prototype system is tailored to GIS-based sampling of biological, chemical and hydrologic parameters within various river cross-sections. Nevertheless, this system is a robust and expandable platform that can address the needs of field workers in a variety of contexts, including research, monitoring, resource management and emergency response. In addition, the value added by the system to off-the-shelf field data collection and mapping tools, such as Esri ArcPad, is substantial.
In this brief paper, we will discuss the development of the various components included in the ENVIT system from an information technology and geographic information system (GIS) perspective and its application to environmental data collection. Section 2 details the specific environmental application for which the system prototype was designed, a watershed quality and quantity study. In Section 3, integration of the computing platform and environmental sensing equipment is discussed. Section 4 details the wireless and web-based technology utilized for transmitting data to and from multiple platforms. Section 5 highlights the use of GIS technologies for mobile and Internet computing utilized in the development of the system. Finally, field deployment and other potential applications are discussed in Sections 6 and 7.
The ENVIT Integrated Field Data Collection System was created in the context of a watershed study designed to gather water quality and water quantity data, as well as geopositional information, at various river cross sections within a river network. The scientific objectives for the field study were to sample water quality parameters that provide an indication of the transport of non-point source pollution (e.g. nutrient loading or bacterial transport) within a watershed impacted by agricultural runoff. Data collection during an intensive field campaign is designed to support hydrologic and transport modeling efforts in the watershed by augmenting available historical data. A GIS-based, semi-distributed, conceptual hydrologic model is currently being calibrated based on the field and historical hydrometeorological and water quality data (i.e. Hydrologic Simulation Program - Fortran with BASINS GIS interface). In addition, the field campaign can be used to validate or verify GIS data layers previously gathered for the watershed and available to field workers on the mobile computers.
The site chosen for carrying out the field testing and data collection campaign was the Williams River watershed in New South Wales, Australia, as illustrated in Figure 1. In addition to confronting water quality issues due to agricultural runoff, the Williams River is an area of active hydrological research and monitoring by various Australian research groups and government agencies (e.g. Wooldridge et al., 2001). Moreover, a network of water quality and flow monitoring stations have been installed in the watershed that are capable of telemetering real-time data and displaying them within a GIS-enabled web browser (HITS, 2002). The availability of this pre-existing infrastructure was an attraction for testing the ENVIT system prototype.
The Williams River catchment, approximately 1260 square kilometers in size, is located within the lower Hunter Valley, about 30 kilometers north of Newcastle. The watershed is characterized by a mountainous, forested region in the north and a series of low, rolling hills, used for cattle grazing, in the central and southern portions. The warm, temperate climate is modulated by an orographic effect imposed by the mountain range and a maritime influence due to its proximity to the southeastern Australian coast. Land-use patterns are typical of this area of Australia, with the majority of the natural eucalyptus forest removed for cattle grazing and agriculture in the lower catchment area. Cattle grazing, in particular, has led to the eutrophication of the Williams River with high phosphorus concentrations observed during storm events (Nolan et al., 1995). Additional problems with algae blooms and low dissolved oxygen concentrations have also been reported. Due to the high value placed on water supply from this watershed, various efforts are currently underway to monitor and model the water quality within the Williams River.
The ENVIT Integrated Field Data Collection System consists of a series of ruggedized mobile computers with the capability of acquiring, storing, displaying and transmitting environmental and geopositional data. The system components are similar to the proposed tele-geomatics conceptual framework outlined by Zingler et al. (1999) for wireless field data collection of environmental observations. Figure 2 illustrates the hardware components for the computing and sensing technologies incorporated into the mobile data collection system.
Conceptually, the system design intends to provide mobile users with a robust and integrated system with the potential to gather data from any digital sensor and log the information onto a pre-configured database on the mobile device. To acquire data, the ENVIT system has interfaces for two different sensors: a GPS sensor and a water quality probe. In addition, manual input of environmental data from other field equipment or observations from the field study are possible through the mobile device software application. To store data, the integrated system contains a mobile device database application which stores records gathered from both the sensors and the user interaction. To display data, the electronic notebook has the capability of uploading records onto Esri ArcPad GIS for cartographic viewing or viewing a GIS .NET web service and an Esri ArcIMS site running on a remote server through the mobile browser. Finally, to transmit data, the ENVIT system has the capability for wirelessly transferring data from individual mobile devices to a centralized laptop which is then relayed wirelessly to a remote web server.
The mobile device application consists of a graphical user interface (GUI), code to support the connection to the database and code to interface both the water quality probe and the GPS sensor. Figure 3 illustrates an example screenshot from the mobile device application. The application is programmed using Microsoft eMbedded Visual Basic 3.0 and runs on the Microsoft Windows CE 3.0 OS, common to many hand-help computers. Scientific and engineering calculations have also been incorporated into the electronic notebook, including a calculator, a hydraulic engineering equation solver, a time-series plotter and an engineering unit converter. These modules allow field users to quickly compute standard discharge rates and geochemical fluxes.
The development of the mobile device application with Microsoft eMbedded Visual Basic resulted from a need to have access to Microsoft Foundation Classes and Active X components. In particular, connections to the mobile database were facilitated within the Visual Basic development environment. Alternatively, consideration was given to implementing the electronic notebook within Esri ArcPad, utilizing the form building capabilities of ArcPad 6.0 Beta Studio. Although powerful, existing limitations with regard to interoperability with other applications prevented the extensive use of ArcPad. Nevertheless, the mapping capabilities in Esri ArcPad were utilized for mobile cartography during the field campaign.
The core application for the ENVIT Integrated Field Data Collection System is a database designed to store and manipulate the data gathered from the manual user input or the automated sensor input. The database is capable of simultaneously updating multiple users working on various platforms. For this purpose, we chose to use Microsoft SQL Server on a field laptop to serve as a central data repository and Microsoft SQL Server CE for each mobile device database. Currently, the database design conforms to the data types, instruments and locations utilized during the field campaign. The design, however, is sufficiently flexible to configure many other data collection processes. Furthermore, a laptop application in Microsoft Visual Basic.Net was developed to serve as a database configurator and facilitate the creation of future applications. Finally, the updating and syncing of the database records occurs wirelessly through the use of wireless cards and an outdoor field router.
Georeferencing the environmental parameters is automated within the system database application, which consists of a master copy operating on a field laptop and local copies on the mobile devices. A key feature is the capability of the master database to simultaneously gather and process newly acquired data on multiple mobile devices. The design is made robust and expandable by abstracting the components of the field data collection process. In general, the database consists of data tables describing the project, location, field operator, equipment and data. The database design permits new sensors or project locations to be integrated as needed.
Coupling a mobile computer with an environmental sensor is a key development of the ENVIT Integrated Field Data Collection System since it provides the capability of data collection for water quantity and quality parameters. For the prototype system, we have developed a manual input interface to the Hydrolab Minisonde multiple parameter water quality sensor and automated control of the Teletype PCMCIA GPS, attached directly to the mobile computer. The water quality probe has the capability of measuring water pH, dissolved oxygen, temperature, conductivity, turbidity and depth, among other parameters. The manual input of water quality data into the mobile device software application occurs through an easy-to-use cross-sectional diagram. For the precise location of the sample in a river cross-section, the user can specify the parameter measured and the value returned by the instrument. In a similar fashion, cross-section based graphical input is supported for flow velocity measurements taken with a digital current meter, chemistry analysis from a field-portable spectrophotometer and biological samples measured for bacterial and algae counts after incubation.
Associating the environmental parameter with a geographical location is a key element of environmental field data collection. The automated GPS sensor onboard the mobile computer logs geoposition continually during the field sampling activities at a sampling location. Positional averaging of the GPS signal is performed to ensure accuracy in the measurements. During a riverine study, the positions of the banks are the most important to obtain accurately. Upon measuring the cross section, this local coordinate system can be transferred to a known geographic location with the bank GPS measurements. The tagging of GPS location to the precise sampling site is achieved in this fashion in order to georeference the environmental measurements.
In order to seamlessly share data amongst scientists and engineers in both the field and at a remote location, the ENVIT Integrated Field Data Collection System includes a wireless local-area communications network. This network transmits data, specifically the SQL database contents, from each mobile device to a field laptop that serves as central data repository. An 802.11b field router with an omnidirectional antenna operating at 2.4 GHz and a series of wireless cards are used to provide a signal radius of several kilometers. This range permits various field teams to work independently at different locations within a watershed and update the central database instantaneously. Figure 4 displays the hardware architecture for the ENVIT system including the wireless and Internet technologies.
While transmission to the field laptop is performed through a wireless local-area network, data transmission to a remote location requires the use of an web-enabled device. Based on wireless coverage from various services providers, the use of a dual-frequency mobile phone was chosen as an appropriate technology. The ENVIT system prototype consists of a mobile phone utilizing a GSM/GPRS service from providers operating at 900 MHz in Australia and 1.9 GHz in Boston. The mobile phone is connected through a serial port to the field laptop and serves to transmit data periodically to a web server located at MIT. Periodic data transmissions are captured by a customized web services application and displayed within a browser for viewing by any interested party. The data display can include both time-series and maps of environmental parameters associated with a location in the field site.
The ENVIT system integrates mobile, desktop and Internet tools for geographical information systems. The use of a GPS sensor and mobile GIS mapping in ArcPad is a powerful combination that allows for the spatial display of geopositional data onto pre-loaded maps of the field site. The system includes a customized version of Esri ArcPad GIS that allows the user to control the GPS, view GIS data layers for the field site and load records that associate the geopositional data with the environmental parameters collected at a specified location. The customization capabilities of ArcPad 6.0 were leveraged for the scripting and user interface elements required to carry out the control of the GPS and the visualization of sampled environmental data in a geospatial reference frame. The final product is a data layer that displays the spatial pattern of an environmental parameter (e.g. pH, turbidity, nutrients, conductivity, dissolved oxygen) and its relation to other watershed characteristics (e.g. topography, land use, soils, streams). Data collection from multiple teams can be visualized together as the field study progresses with more powerful geospatial processing occurring within a desktop GIS loaded onto the field laptop server.
The distribution of geospatial data is enhanced in the ENVIT system with the use of two GIS web services applications residing on a remote web server at MIT. Periodic uploading of field data, in the form of the database contents, is processed by a .NET webservice that can display the field samples overlaid onto a watershed digital elevation model (DEM) and stream network map. The custom-tailored application provided a flexible tool for displaying the field data in near real-time for distribution on the Internet. The mobile web browsers used by each team could also view the website using the wireless local-area network. In this way, field personnel had two method of accessing the data collected and integrated from multiple field teams. In addition, a post-field study effort resulted in the development of an Esri ArcIMS web site for the display and distribution of GIS data layers collected for the Williams River prior and during the field campaign.
The deployment of the ENVIT Integrated Field Data Collection System is a coordinated effort between various field teams and individuals with specific data gathering tasks. For the watershed study in the Williams River, three field teams consisting of six to seven individuals were assembled. Each team was equipped with two ruggedized hand-held computers, a wireless card, water quality and GPS sensors, a streamflow gauge, a lightweight battery pack and other chemistry and biology measurement kits. Each mobile device either operates the two sensors, or is used for wireless transmission and manual data input ( e.g. streamflow, chemistry, biology). The integrated infrared port is used to transmit data between the two hand-held computers. During the field campaign, the three field teams gathered geopositional and environmental data at three separate river cross sections and transmitted data to a single station equipped with the wireless router and the field laptop. From this station, periodic data transmission was made to the web server at MIT and processed data displayed back to the field teams through a web browser.
In order to ensure appropriate field deployment of the technology, the ENVIT system also includes two hardware components that provide waterproofing and extend battery life. Modifications to an existing hand-held case were made so that the mobile devices, GPS and related cabling are protected from dust and accidental submergence. In addition, custom power management circuitry was designed and built to power both the mobile devices and the water quality probes from a single power source. This power pack enables approximately forty hours of continuous system use. These hardware developments are critical for field deployment since the mobile devices are susceptible to water damage and have high power consumption rates.
The design of the ENVIT Integrated Field Data Collection System has focused on creating a robust system capable of gathering environmental and geopositional data from selected devices. Although the prototype is specific to a water quality and quantity study, the system design is sufficiently flexible to add other environmental sensors with ease. Only minor changes to the user interface and the creation of device drivers are required. Furthermore, the database, wireless and web technologies are transparent to a new instrument.
A potential application for the system is the response to water-borne chemical or biological disasters. The successful field scale testing has demonstrated the system’s capability to serve in situations that require a coordinated response by multiple field teams. The advantages this system are evident in emergency response situations that involve: (1) a critical time constraint; (2) an expansive study area; (3) computationally extensive analysis or modeling; (4) remotely located decision support and management; and (5) multiple individuals requiring processed field data.
A second application area is the use of in-situ measurements for the verification of remote sensing data from terrestrial or satellite platforms. Deployment of multiple teams measuring ground-based environmental parameters over the sensor footprint can provide a mechanism for remote sensor calibration and validation during intensive field campaigns.
In this paper, we present the development of an integrated data collection system designed for acquiring, storing, displaying and transmitting environmental and geopositional data during field campaigns. The project goal is to provide a more accurate, efficient and robust method for gathering data by integrating existing hardware and software components with new mobile, wireless and Internet technologies. Field testing of the integrated system has been within the context of a water quality and streamflow study in the Williams River watershed located in New South Wales, Australia. Through field deployment, the full capabilities of the integrated system were demonstrated. Results from the field study suggest that the prototype system is most useful for time-critical, intensive field campaigns carried out by multiple research teams. Furthermore, the use of distributed wireless sensors was shown to be a viable tool in environmental field research. Mobile and Internet mapping are useful for summarizing and displaying field collected data during extensive sampling studies.
Briner, A.P., Kroneberg, H., Mazurek, M., Horn, H., Engi, M., Peters, T., 1999. FieldBook and GeoDatabase: tools for field data acquisition and analysis. Computers and Geosciences. 25, 1101-1111.
Carver, S., Heywood, I., Cornelius, S. and Sear, D., 1995. Evaluating field-based GIS for environmental characterization, modelling and decision support, Int. J. Geo. Inf. Sys., 9(4): 475-486.
Hunter River Integrated Telemeterying System, 2002, Available at: http://hits.nsw.gov.au
Nolan, A.L., Lawrance, G.A. and Maeder, M., 1995. Phosphorus speciation in the Williams River, New South Wales: Eutrophication and a chemometric analysis of relationships with other water quality parameters. Marine and Freshwater Research, 46: 1055-1064.
Pundt, H. and Brinkkotter-Runde, K., 2000. Visualization of spatial data for field base GIS, Computers and Geosciences, 26: 51-56.
Vieux, B., 1991. Geographic information systems and non-point source water quality and quantity modeling. Hydrological Processes. 5: 101-113.
Wilson, J. D., 2001. The next frontier: GIS empowers a new generation of mobile solutions, GEOWorld, 6: 36-40.
Wooldridge, S., Kalma, J., Kuczera, G., 2001. Parameterization of a simple semi-distributed model for assessing the impact of land-use on hydrological response. Journal of Hydrology, 254: 16-32.
Yang, M-D., Merry, C.J. and Sykes, R.M., 1999. Integration of water quality modeling, remote sensing and GIS. Journal of the American Water Resources Association. 35(2): 253-263.
Zingler, M., Fischer, P., Lichtenegger, J., 1999. Wireless field data collection and EO-GIS-GPS integration. Computers, Environment and Urban Systems, 23, 305-313.
The development of the prototype system was carried out as part of a MIT/Microsoft Alliance-sponsored I-Campus project by the ENVIT Student Group, an MIT-recognized student organization. The authors led a group of undergraduate and graduate civil and environmental engineering students in the development and application of the system. We gratefully acknowledge the support of the students involved in the project and field testing: Neeraj Agarwal, James Brady, Nancy Choi, Chrissy Dobson, Arthur Fitzmaurice, Eric Lau, Anna Leos-Urbel, Linda Liang, Rose Liu, Kan Liu, Brian Loux, Aurora Kagawa, Kris Kolodziej, Patricia McAndrew, Kevin Richards, Laura Rubiano Gomez, Kim Schwing, Russell Spieler, Chin-Huei Tsou, Lisa Walters, Amy Watson and Keyuan Xu.
Enrique R. Vivoni
Graduate Research Assistant, Ralph M. Parsons Laboratory
Massachusetts Institute of Technology, Cambridge, MA 02139
vivoni@mit.edu, (617) 253-1969
Richard Camilli
Graduate Research Assistant, Ralph M. Parsons Laboratory
Massachusetts Institute of Technology, Cambridge, MA 02139
romeo@mit.edu, (617) 258-9534
Mario A. Rodriguez
Software Engineer
Rovia, Inc., Cambridge, MA 02139
mario@rovia.com, (617) 497-7850
Daniel D. Sheehan
GIS Specialist, MIT Information Systems
Massachusetts Institute of Technology, Cambridge, MA 02139
dsheehan@mit.edu, (617) 252-1475
Dara Entekhabi
Professor, Ralph M. Parsons Laboratory
Massachusetts Institute of Technology, Cambridge, MA 02139
darae@mit.edu, (617) 253-9698
Sheila Frankel
Assistant Directory, Ralph M. Parsons Laboratory
Massachusetts Institute of Technology, Cambridge, MA 02139
sfrankel@mit.edu, (617) 253-2339