Water-flow in soil
What do you think of when I say: “green roof soil”?
If you are a green roof engineer, your thoughts will likely wander to the FLL guidelines, gravel, and “dirt needed for plants to grow.” Of course, you are right in this, but the soil is also one heck of a complicated thing, and even small tweaks in the media used on a green roof can have enormous impacts on its functionality.
One thing that does change when the soil composition is altered is how water moves through the soil. The impact of water movements through this shallow, often merely 6’’ (10 cm) layer, might seem insignificant. Even trivial. But…
…over the years we have started to understand that this is not the case. A deeper understanding of the hydrological behaviors of green roof soil media might well be the key to further advancements of soils developed for local climates. Soils that are matched to vegetation, precipitation, and temperature, as well as the other engineered materials used in the construction. On a local level. This harmony results in maximized retention, plant happiness and lowered nutrient runoff.
But, let’s start with basic water movements - how does water move through the soil?
Water flow in soil
Water likes water like Kanye loves Kanye. Have you ever watched the droplets on a cold bottle of beer, or on a car window on a rainy day, as they travel downwards, colliding and fusing with other droplets in their paths?
It almost looks as if they were drawn to each other. And that is precisely what they are! The reason for this lies in the polarity of the water molecule: water attracts water.
This happens in soil too.
Preferential flow vs. matrix flow
Sometimes, and too often in my humble opinion, green roof soil is seen as something homogenous and simple: dirt that sucks up water like a sponge with the fluid evenly distributed through the media. This is, however, far from the truth. Just as with the droplets on the cold beer bottle, water meets water in the soil and follow specific paths.
There are two main processes that describe water movements through soils.
The first process is called matrix flow. This is a rather slow process in which water gradually fills up the pore spaces of the media. The pores are not filled up simultaneously but with different ease depending on the surface adsorption properties of the soil particles. Water sticks differentially well to different surfaces.
Let’s take an example of how water can get adsorbed or rejected. Some surfaces simply hate water. I have a Labrador retriever, and she loves swimming in icy waters. The cold wetness doesn’t bother her the slightest. Her fur is full of hydrophobic (“water-hating”) fats that repel the water. Her skin never gets wet, so she doesn’t freeze.
So-called hydrophilic surfaces adsorb water. My cotton trousers, on the other hand, adsorb, and absorb, all the ice-cold water my dog shakes off after her bath.
The second soil water flow-process is called preferential flow. This process displays highly enhanced flux and often only represent a small fraction of the total media. The paths often follow cracks found in the porous media, but also worm- and root holes become preferential paths, showing how vegetation and fauna play a central role in how water moves through soils. Also here, particle hydrophobic or hydrophilic properties play a role.
These preferential paths often create something called “fingers,” which are formations in which water flows fast and deep into the soil. Finger formation is generally a healthy thing in normal soils as it creates a good soil water/gas balance. Dry patches mixed with wet ones mean that root systems do not completely get flooded nor do they lose out on water. Flooded roots can result in lower root respiration and, in the end, the root may rot and die. Thus, a good soil balance between air and moisture is important.
The simplest infiltration models merely assume that soil consists of two homogenous layers and that there is a wetting front parallel to the soil surface. Gravity and matrix suction cause the wetting front to move downward. This is, of course, a gross over-simplification, as I have already mentioned.
Finger forms when the so-called wetting front becomes unstable, caused by an instability of the infiltration flow. They can form during heavy rainfalls. These fingers move down fast and bypass the unsaturated soil medium creating preferential flow paths.
There is a vast difference in how matrix and preferential flow affect the movement of nutrients, salts, and pesticides through the soil.
Think about how water moves through the soil when
Thus, if you have supplied nutrients on top of the soil, these may be flushed out through preferential flow paths without doing much good at all. Several studies have shown that deeply infiltrated water that followed formed fingers had similar nutrient and salt concentrations as the surface soil layer, whereas water following the matrix pathway showed solute concentrations that were similar to the drain.
Thus, not only water leaves the roof fast, but also nutrients quickly exits soils with strong preferential flow paths. It is simply flushed out. This loss of nutrients may require more frequent roof fertilization, something that is not only bad for the environment, but also for the
We need better knowledge of soil structure and
We know way too little about how water moves through green roof media, and we often make broad assumptions extrapolating from systems that have little to do with green roofs.
However, hopefully, we will soon see more natural new green roof soil media with thriving soil biology hit the markets. I’m excited to find out how these soils behave regarding water and nutrient flow.
We truly have exciting times ahead!
Don’t hesitate to contact our experts if you have any questions or proposals. We are always open to new ideas!
Reading tip: water-flow in soils
Ritsema, C. J., Dekker, L. W., Van den Elsen, E. G., Oostindiel, K., Steenhuis, T. S., & Nieber, J. L. (1997). Recurring fingered flow pathways in a water repellent sandy field soil. Hydrology and Earth System Sciences Discussions, 1(4), 777-786.