Subsurface Hydrological Monitoring of a Watershed with Hybrid Sensor Networks


This abstract describes novel instrumentation for in situ hydrological monitoring of watersheds. Our system autonomously measures various attributes of the watershed soil, including chemical composition, moisture, temperature, and resistivity. The measurements are taken at several depths and are communicated to a processing server over the existing GSM cellular infrastructure. Existing hydrological methods suffer from shortcomings in accuracy, resolution, and scalability. Their fragility, high power consumption, and lack of long-range communication capability necessitate frequent site visits by experts. Cabling requirements and large size limit their scalability and make such systems prohibitively expensive. Our system is comprised of a network of sensor strings, each of which connects up to 100 sensing nodes on a communication line that can be up to 100m in length. Each of the nodes is comprised of the sensors needed for measuring soil attributes of interest, as well as a microcontroller with basic communication and processing capabilities. A relay point at the surface aggregates data from the nodes and wirelessly transmits it to a base station that serves as a gateway to the outside world. The base station aggregates data from multiple strings and utilizes the GSM cellular infrastructure to communicate the data to a data collection server, and to receive remote updates to be downloaded to the sensor strings. Ultra-low power design and remote maintenance result in an unattended field life of over 5 years. The system is scalable in area and sensor modality, as covering a larger area would only entail additional sensor strings, and the nodes are designed to facilitate the interfacing of additional sensors. The system is robust, as the only exposed portion is the relay point. Data collection and transmission can be event-driven or time-driven. Battery power, which can be supplemented with solar harvesting, and wireless short- and long-range communication, eliminate the need for surface wiring and significantly reduce the cost of system deployment. Currently, our estimate is a cost of less than $30 for each sensor string for small batch production, which compares very favorably to existing systems that have limited capabilities yet cost tens of thousands of dollars. In summary, we believe that the proposed system has the potential to significantly improve hydrological monitoring. The system enables the collection of data at a scale and resolution that is orders of magnitude greater than any existing method, while dramatically reducing the cost of monitoring. The quality and sheer volume of information collected as a result will enable previously infeasible research in hydrology.


Electrical and Computer Engineering

Second Department

Geosciences and Geological and Petroleum Engineering

Keywords and Phrases

Aggregates; Batch Production; Communication Systems; Cost Engineering; Electric Power Generation; Hydrology; Infrastructure; Instrumentation; Networks; Servers; Soil (Material); Watersheds; Wireless Communication

Document Type

Article - Conference proceedings

Document Version


File Type





© 2007 American Geophysical Union (AGU), All rights reserved.

This document is currently not available here.