The Coca-Cola Company with its self-righteous social responsibility program wants to become water-neutral, i.e. they want to return to the environment all the water they use to make beverages, wash bottles etc. So, they have funded the WWF to plant trees. How much water does planting trees save? Well, that was our job. When you have trees, the rainfall does not hit the soil directly, and instead trickles down into the soil and gets absorbed into it. With trees you have a different soil composition and different shrubs and bushes as undergrowth, both of which effect soil porosity. The more porous the soil the more water it can absorb. Absorbed water is slowly released into underground springs, which are a vital source of water for the indigenous population as well as the animals and plants. Water that does not get absorbed runs off the side of the mountain and flows into rivers. This water is considered wasted.
Infiltration, is water seeping into the soil, and we were measuring how its rate is affected by soil composition, slope, canopy cover, tree number and species, shrub and grass density. We visited 18 sites spread over different intensity forest areas. How is it measured? Why using an infiltrometer of course! This fancy name is just two metal tubes, one wider than the other. Both are hammered into the soil concentrically. The outer wide ring is to prevent lateral spread of water. Both rings are filled with 4 cm (or whatever else is suitable) of water. Water slowly seeps into the soil from both rings. Once saturated the rate becomes constant and this is your infiltration rate.
We planned to apply a simple linear regression model onto our data hoping for a wonderful magical linear fit. Unfortunately the stepwisefit (Matlab linear regression function) thinks only canopy cover affects the infiltration! A lot of serious genius is about to go into this project to make the Math work.