Sensing Soil Moisture with Microwaves - How Does it Work?

Maik Wolleben

Oct 15, 2021

In this post, I talk about Skaha's microwave soil moisture sensor. But first, I should clarify what I mean when I talk about microwaves. Microwave radiation is electromagnetic radiation at radio frequencies, the part of the electromagnetic spectrum that your cell phone uses (700 MHz, up to several GHz) or when you tune your radio to your preferred FM radio station (98.5 MHz, for instance), then microwave radiation is used to transmit speech, music, videos, etc. In both examples, there is a receiver and a transmitter, such as a cell tower or a radio broadcast antenna. Skaha's soil moisture sensor, on the other hand, is only a receiver. So, how can we measure soil moisture passively, without using a transmitter?

Well, the soil itself is the transmitter! Let's see how this is possible. Imagine being a blacksmith and heating a metal rod to 500 degrees Celsius or more. If it's dark enough in the workshop, you will see it glow. First red, then yellow, depending on how hot the metal becomes.

The light you see is emitted by the atoms and molecules that make up the metal rod. You may remember from school that heat is nothing but the jitter and motion of the elementary particles. For example, air molecules (O2, N2, etc...) vibrate just a little more during a hot Okanagan summer day than a cold Alberta winter night. Particles that vibrate lose their energy by emitting electromagnetic radiation. In the case of the metal rod, some of that electromagnetic radiation happens to be in the visual part of the spectrum. This is what makes the metal glow when it's hot.

So what does that have to do with soil moisture and Skaha’s sensor? Many materials emit electromagnetic radiation, even when they are not glowing hot. Soil at ambient temperature, for example. This is the so-called thermal radiation or black body radiation, naturally generated by materials. However, we can't see this radiation with our eyes because the wavelength of said radiation is in the microwave range, not the spectrum's visible part. Fortunately, we can use sensitive radio receivers to measure its intensity, which is precisely what all our microwave soil moisture sensors do.

But we are not quite there yet. We also have to look at the mechanism that modulates the soil's radio emission. Without it, all soils would have the same level of microwave emissivity, which would be terribly dull. The most critical ingredient that affects microwave emission is water. Water is a killer for microwaves. Microwaves can't travel in water; they are either reflected at the surface or absorbed once inside. And that is how we can distinguish wet soil from dry soil. Dry soil will "glow" just a little brighter at microwave frequencies. Its microwave emissivity will be higher than moist soil because the water in wet soil reflects and absorbs the soil's thermal emission. This is important: The more humid the ground, the less microwave radiation makes it to the surface. This change of emissivity at the surface is what we can use to determine soil moisture.

The nice thing about microwaves is that they penetrate materials, which is why you can use cell phones inside buildings because walls are transparent to microwaves (well, to some extent). This also allows us to sense soil moisture within the soil and not just at the surface. But it depends on frequency. The lower the frequency, the deeper the sensing depth. The Radarsat satellite, for example, operates at 5.3 GHz and measures moisture in the top few cm. SMAP, another soil moisture satellite, works at 1.4 GHz and senses roughly the top 5 cm. Our system works at 450 MHz and we can sense down deep into the root zone of most crops. Our sensing depth is about half a meter (20 inches) and even deeper in dry soils.

So, how accurate is it? This is surprisingly difficult to answer and I will get into this in one of my next blog posts. But as an example, soil moisture sensing satellites such as SMOS or SMAP usually aim for an average accuracy of 0.04 m3/m3, which means +/- 4% volumetric water content. Skaha's sensors have similar accuracy. Compared to in-situ soil moisture probes, which usually get you within +/- 1% or better, the accuracy of microwave sensors is not as good. However, remote sensing allows you to map large areas without ever physically touching the soil. That's obviously an advantage if the goal is to map soil moisture over hundreds or thousands of acres in an efficient manner.

The field above is an example where the soil moisture (red for dry soil and blue for wet) follows topography really well. The elevation scale is exaggerated. The data were collected with a pair of our soil moisture sensors mounted to a pickup truck.

With all this theory behind us, how exactly does Skaha turn radio waves into soil moisture maps in practice? Usually, one or two microwave sensors are attached to an irrigation pivot or a vehicle such as a quad, pickup truck, or seeder and driven along parallel transects at a speed of 35 km/h or less. Data are collected by our telemetry unit and uploaded to our server via cellular uplink. On our server, we convert microwave emission into maps of volumetric water content. If we know soil texture, we also produce maps of plant-available water.