Modelling stream shade: 1. Verifying numerical simulations with measurements on simple physical models

Sunlight exposure of streams controls their thermal regime and light availability for photosynthesis. Restoring riparian shade to formerly forested streams in cleared land is an important step in the restoration of stream ecological health. Methods exist for measuring existing stream shade, but management tools are needed for predicting future shade as riparian plantings grow. We refine an existing geometrically-based numerical model for diffuse lighting which can be linked to models of tree growth to investigate how shade changes over time. The model predicts 'diffuse non-interceptance' (difn) which is an index of long-term averaged lighting. Shade of real (meandering) streams is bounded by two computationally straightforward cases: a perfectly straight channel (canyon model) and a channel meandering so tightly that its shade geometry 'collapses' to that of a circular pool (cylinder model). In this paper, we verify model results against measurements of shading made with a canopy analyzer instrument in precisely-constructed physical models of both the cylinder and canyon cases. For both canyon and cylinder geometries the observed and predicted difn matched closely (over a wide range of the ratio tree height to stream width, h/w) corroborating the numerical model. As expected, light exposure at the channel mid-point was higher than at channel edges, highlighting the need to consider variations in lighting across the channel when predicting water temperature and aquatic plant growth. Over the range of h/w where difn changes rapidly, shade was appreciably higher for the cylinder than the canyon case.