During the course of the week, we have been constantly increasing the wave conditions that our plants get exposed to: the last wave spectrum of the week finally also contained some waves that spectacularly broke and washed white water across our setup. We were surprised how little biomass came floating up to the surface, but now cannot wait to see what these conditions did to the plants and soil.
The SET measurements that we perform (see previous blog entry) are crucial for documenting soil elevation changes, but they only give us data at certain points within each pallet. To complement these, we are using structure from motion (SFM) to produce 3-dimensional models of the pallets for every day. SFM requires us to take lots of photos (about 200 per row of pallets), from lots of angles, ensuring that we have plenty of overlap between them. The software then identifies matching points in multiple photographs and triangulates their positions in space to build very high-resolution (mm) 3D models. We can then use these models, combined with ground control points, to ensure accuracy, to look at changes in soil volume across entire pallets and to investigate how erosional features change over time in their geometry. We can validate these measurements against the SET measurements amongst others. The SFM approach isn’t only useful for the soil surface, we can also use it to look at plant bio-volumes and characteristics. Check out how clearly you can see the individual plants in the model and how many photos were necessary to produce this model: