Projects Nordbalt-Ecosafe Sensor monitoring

Sensor monitoring

What are sensors?

Monitoring is an important part of any study of eutrophication, and in NORDBALT ECOSAFE we are exploring the use of the relatively novel method of monitoring by sensors. A water quality monitoring sensor is a device that can monitor a set of parameters frequently and in real-time. The sensors can be placed in streams, lakes, groundwater wells, or coastal waters. In our policy brief we list advantages and challenges with sensor monitoring and give examples of sensor use. Note that the brief is produced in several languages.

Real-time data from our interactive map

One of our aims is to enhance the citizen awareness on water quality matters by offering real-time access to a set of sensor data in stream monitoring stations. Check out our interactive map: Where you find a yellow star, you may click on that star, and you will be taken to a site showing real-time data from that stream. You can also access the data from links in our fact sheets, found below.

Watch up: Our real-time data are not yet quality controlled

It is very important to understand that on-line data have not yet been quality controlled. This means, that before we use the data for research or management purposes, we have to go through them and ‘clean up’ any incorrect or suspicious data. Such errors can occur for many reasons, some of these are listed below:

  • Power outage or other problems with the equipment. 
  • Leaves or other organic debris may temporarily clog the lens, giving short-term artificially high levels.
  • Biofilm or sediments on the lens may give longer periods of too high recordings. (this is often detected when the sensors are cleaned manually).
  • Some sensors can only be operated in the time of year without ice and snow.

The figure below shows some examples of such typical errors. Commonly, we keep the original data series, but replace it with one where the suspicious data have been removed, unless we can find an explanation of these data. Despite of such potential errors in the sensor dataseries, we believe that displaying real-time data to citizens will be of value, also because the local inhabitants may have other explanations to some of these anomalies.

NORDBALT ECOSAFE is working towards common procedures for quality assuring sensor data, and a PhD student at AU is presently writing a paper on an automated quality control tool. The tool is expected to be available on-line, in open access, later this year. The developments will be posted at this web site. 
The figure illustrates typical quality issues with a turbidity sensor. Here, turbidity (orange graph) and water level (blue graph) are recorded each 30 minutes in River Kråkstadelva, Norway. Blue stars: Illustrate turbidity peaks without corresponding water level increases: These can be true increases, or caused by, e.g., organic debris that shortly block the lens. Green square: Period where the turbidity sensor was not working (power outage). Red circles: Assumed too high levels, possibly due to biofilm or sediments clogging the lens for longer periods, before the sensor was manually cleaned, with an obvious drop in NTU after cleaning.

Link to sensor stations in Denmark

Link to sensor stations in Norway

Skuterudbekken

  • Station name: Skuterud
    • Link to on-line turbidity (FNU) (30 days) https://kart8.nibio.no/nordbalt_ecosafe/sku.html
    • Coordinates: 59°41′6″ N 10°49′52″E
    • Catchment area: 4.5 km2
    • Land use: agricultural land 61%, forest 29%, urban areas 8%, wetland 2%
    • Dominant soil type: silty clay 76%, silty sand 15%, sandy clay 7%
    • SEBA sensor
    • Please note: Sensor data are updated each day at 9:00.
  • Station name: Kråkstadelva

Link to sensor stations in Sweden

Station name: Hågaån

  • Link to online turbidity (NTU), depth, conductivity and water temperature (30 days)
  • https://ghapps.slu.se/plotapp/
  • Coordinates: 59.806178, 17.604204
  • Catchment area: 122 km2
  • Land use: forest 41 %, agricultural land 25%, urban areas 18%, open land 12%, wetland and lakes 4%
  • Soil type: clay 28 %, moraine 19%, rock and thin soils 19%, silty clay 15 %, hard surfaces 6%, peat 5%, 8 % other
  • Type of sensor: InSitu Aquatroll 3000
  • Deployed since 2018, all year deployment
  • Please note: Sensor data is updated every day at 06.00

Link to sensor stations in Finland

Station name: Lepsämänjoki

Station name: Leppioja

Different sensor data

Policy Brief on sensor monitoring

Read our 4-page brief with practical information on the use of sensor monitoring, available in the following languages:

Revised 27.02.2024
-
Else Vihlborg Staalsen