Environmental sensing with wireless networks

Collecting multiple data on a limited budget can be a challenge. NIWA's Instrument Systems group uses its own product, SMARTi, an inexpensive local wireless interface, in a practical low-cost radio network that minimizes installation costs and system operating charges.

The network collects data from 46 discrete sensors, distributed over two hillsides, as part of a water catchment study. There are 32 soil moisture/temperature sensors and 14 water level sensors located in groundwater bores.

Sometimes it is necessary to record data from several monitoring sites, each within sight of each other, yet too far apart to easily connect by wire. While each site could have its own data logger, which would independently send data to a database via a cellular or satellite network, the on-going charges may become prohibitive with increasing numbers.

SMARTi has a lower initial cost and only a fraction of the on-going costs of alternative options.

Why so many data?

Water is an invaluable resource, but it can also be a hazard if we get too much of it in the wrong place at the wrong time. Reliable prediction of droughts and flooding is becoming increasingly important. Quality hydrological data is essential to develop and fine tune the hydrological models that predict the consequences of rainfall events.

The purpose of the intensive monitoring included in the Waipara Project is to develop a view of surface and ground water together, as linked resources. Headwater catchments, such as the Waipara, play an important role in how rainfall is routed as runoff and stored as groundwater. This project is aimed at understanding how a catchment's soils, landscape, slopes and vegetation, partition rainfall into surface flow and groundwater flow.

This dense network of sensors, along with a local compact weather station, allows us to measure and monitor soil moisture, groundwater recharge and rainfall. We are also monitoring surface flows at the bottom of the gully, using a telemetered water level instrument at a V-notch weir. By combining these, over time, we can estimate the catchment water balance, which is essential in partitioning rainfall runoff from recharge. This in turn allows us to estimate the time it takes rainfall, at headwater catchments, to reach downstream catchments and rivers, via streams and groundwater.

The challenge

Because of the required density of monitoring, we needed an inexpensive and flexible way to collect the data from each of 16 monitoring sites, each with two soil moisture/temperature sensors (at depths of 30cm and 60cm) and 14 with a groundwater sensor. The instrumentation was to be powered by a battery charged with a small solar panel.

Connecting the sites with wire wasn't practical. Most cable runs would have exceeded the maximum working distance of the electronic devices and laying cables would be time-consuming, especially if buried to protect them from being damaged by animals and farm machinery.

Connecting each of the sensors at each site to its own IP data terminal would increase the initial cost and incur extra on-going network data charges which would be significant as there is no cellular coverage in the area and satellite communications would be required. 

The solution

The best approach we found is to connect wirelessly. We created a wireless sensor network using NIWA's own product, the inexpensive, license-free Smart Radio Interface (SMARTi).

A SMARTi at each of the sixteen remote sites reads the soil moisture/temperature sensors and water level sensors every 15 minutes. The SMARTi processes and transmits the data to another SMARTi located at one of the two base stations.

SMARTi has a low-power mode, where it 'sleeps' much of the time and periodically wakes up just long enough to enable a measurement and to transmit its data to the base station. Because it is off most of the time, it uses very little power and places little demand on the batteries and charging capacity.

Each of the remote locations is within line-of-sight of the base stations it communicates with, this is important to ensure reliable communications. As the sensors are arrayed on north-facing and south-facing hillsides this was easy to arrange as two separate groups, each reporting to its own base station.

SMARTi has a range of up to 1.5 km, and as our distances were less, there is sufficient signal strength to enable good communications. 

The implementation

We installed a group of sensors on each of the two hillsides.

A nearby existing Compact Weather Station (CWS) was within radio range of the north-facing group but we needed to add a Neon Base, or hub, to retrieve the data from the south-facing group. 

Conclusion

This approach to retrieving data from a number of close, but physically-separated environmental monitoring stations can be an inexpensive alternative to individually connecting each site to the server with its own data terminal and incurring extra on-going charges for longer term projects.

NIWA's Instrument Systems Group can look at your monitoring needs and advise on the most cost-effective approach, depending on scale and terrain.

Related articles

Take a closer look at SMARTi and find out what else it can do

Contacts

Instrument Systems: jeremy.bulleid@niwa.co.nz

Science: ms.srinivasan@niwa.co.nz 

This diagram shows how the sensor data is transferred from each monitoring site to a CWS then uploaded to the Neon Server via satellite. The remote wireless interfaces are enclosed in sealed tubes (bottom left) and the soil moisture and water level sensors are wired directly to them.
Monitoring sites being prepared on a hillside as part of a water catchment study. The NIWA Compact Weather Station (CWS) in the foreground will collect the data and send it to the Neon Server via satellite. [Andrew Starr, NIWA]
This diagram shows the overview of the sensor network superimposed on a satellite shot of Langs Gulley. [Satellite photo: Google Earth]
NIWA's SMARTi. [NIWA]
Research subject: Instrumentation