In this new GRL publication, Octavia Crompton explores different ideas of connectivity in drylands, and how they relate to each other.
Hydrological connectivity measures how easily water can move between separated areas. It's often summarised as a distance (L) that measures how far apart connected areas are.
Connectivity determines flooding, risks of erosion and land degradation.
We wondered how connectivity worked when you considered infiltration. The distance that water can flow (Lesc) might be larger than the distance the average water molecule actually traveled before it infiltrated (Linfl). We'd played with this idea for buoyant particles like seeds before, but Tavia built a new infiltrating model to test it for water.
Among other things, this enabled her to make gorgeous videos - here the green surface shows vegetation (which increases infiltration rates), while the slopes show water depth and velocity during the storm. The hillslope hydrograph evolves at the far right.
Tavia found power-law relations between the connectivity metrics and the hillslope runoff coefficient C (relevant to flooding). The relationship changed with storm properties for infiltration connectivity (Linfl) but not for hillslope-scale connectivity (Lesc).
The differences between in the two connectivity metrics are due to the tortuosity (wiggliness) of flow paths. The wigglier the flow paths (the larger the \eta metric below), the greater the difference - presumably because wigglier flow paths cover less ground on the hillslope while giving water more opportunities to infiltrate. This behaviour could be captured at landscape scales with a vegetation Reynolds Number (right hand plot).
What does this all mean?
One idea relates to global versus local connectivity. High global connectivity means resources get lost from the landscape. High local connectivity means resources get moved around within the landscape - which is often essential for ecosystems to function well. What Tavia's work shows is that these ideas are connected, so that redistribution of resources (high Linfl) will always "flirt" with the risk that those resources will be exported - highlighting (again) precarity of drylands and other hydrologically-mediated ecosystems.