Soil Moisture Monitoring Station

When you hear 'soil moisture monitoring station', the first thing that comes to mind is just a sensor in the ground and some numbers on the screen. But in reality, this is a whole system where an error of a couple of percent of humidity can mean either excessive water consumption on hectares or loss of yield. Many people still think that the main thing is to buy the “most accurate” sensor, and the rest is a matter of technology. But I would argue: even with expensive equipment without proper integration into agricultural processes, there will be little use.

How we came to the need for monitoring

I remember one of our first projects at Shandong Linyao Intelligent Agriculture Technology LLC was the installation of a drip irrigation system in a greenhouse near Voronezh. The customer then saved on sensors, deciding that the agronomist would “by eye” determine when to water. A month later we saw that in some areas there was waterlogging, the roots began to rot, in others the plants had slowed down in growth. It was then that it became clear: without objective data on soil moisture, even the most advanced irrigation system is working blindly.

We started testing different sensors - tensiometers, capacitive, resistive. It turned out that for clay soils, for example, tensiometers show a delay of a day, while capacitive ones require frequent calibration. We had to combine them: capacitive sensors were installed at a depth of up to 30 cm, and tensiometers below. This is not according to textbooks, but in practice it gave stable data.

Now that we are designing systems forsmart agricultural parks, we immediately lay down multi-level monitoring. Not only humidity, but also soil temperature and electrical conductivity. Otherwise, you may miss the moment when the root system stops absorbing moisture even if there is sufficient quantity.

Technical nuances that are rarely written about in specifications

One common problem is the so-called 'salt bridge effect'. In the arid regions where we worked, for example in the Stavropol Territory, salts from fertilizers eventually form a conductive layer around the sensor. As a resultsoil moisture monitoring stationbegins to underestimate readings. You have to either service the sensors more often or place them at a slight angle so that rainwater drains and partially washes away the salts.

Another subtlety is the location of the sensors relative to the irrigation system. If we place it too close to the dropper, we will get overestimated values; if too far away, we will get underestimated values. We usually place several sensors in the area where the moisture spreads from the emitter and output a weighted average. This requires more equipment, but the data is representative.

Another point is calibration for different types of soil. Manufacturers usually give universal curves, but they work differently on sandy and loamy soils. For every major project, we do field calibration: we take soil samples, dry them in the laboratory, and compare them with sensor readings. It is labor-intensive, but without this the error can reach 10-15%.

Integration with other systems - where the difficulties lie

When we were developing a project forhigh quality agricultural fieldsin the Rostov region, faced the problem of protocol compatibility.Monitoring stationfrom one manufacturer, irrigation controllers from another, analysis software from a third. I had to write an intermediate gateway that converted the data into a single format. Now at Shandong Linyao Intelligent Agriculture Technology Co.,Ltd we try to use open API equipment even if it is a little more expensive.

An important aspect is the power supply to remote stations. Solar panels are not always a reliable solution, especially during cloudy periods. In one of the farms we had to combine: solar panels + low-power wind generator + backup batteries. This increased the cost of the system, but ensured continuous monitoring throughout the season.

Data transfer is also not easy. In field conditions, GSM communication is unstable, satellite channels are expensive. We often use radio modems in the 868 MHz band, but this requires careful consideration of the terrain. In hilly areas you have to install repeaters, which complicates the infrastructure.

Case studies - what worked and what didn't

Last year, we implemented a monitoring system in the horticultural sector of the Krasnodar Territory. The task there was not only to control humidity, but also to predict watering taking into account weather conditions. We have connectedsoil moisture monitoring stationsto weather sensors and developed an algorithm that took into account evaporation. As a result, the farm reduced water consumption by 22% without loss of yield.

But there was a bad experience with a project in Tatarstan. There we installed sensors only in the top layer of soil (0-50 cm); we did not take into account that fruit trees have a root system reaching 1.5 meters. During the dry period, the trees began to experience moisture deficiency, although according to the station’s readings everything was normal. We had to reinstall the sensors at different depths - additional costs could have been avoided.

Now on the website https://www.lyzhihuinongye.ru we post recommendations on the depth of installation of sensors for different crops. This is precisely the result of that bad experience - it is better to share mistakes than to repeat them.

Prospects for technology development

Now we are testing a system where data frommonitoring stationsare fed into a neural network, which analyzes not only the current humidity, but also the dynamics of its changes, and predicts the optimal watering time. For now, the algorithm requires improvement - sometimes it produces strange recommendations when the weather changes abruptly. But the direction is promising.

Another interesting direction is multispectral cameras combined with soil moisture data. Allows you to identify areas with water stress in plants even before it becomes visible to the naked eye. In one of the smart agricultural park projects, this combination helped to detect drainage problems in certain areas in time.

I think in the coming yearssoil moisture monitoring stationswill become an integral part of not only large farms, but also medium-sized farm enterprises. The main thing is that technologies remain practical and profitable, and not just 'smart toys'. At Shandong Linyao Intelligent Agriculture Technology LLC, we are working on exactly this - connecting advanced developments with the real needs of agricultural producers.