Premium
Using multiple scales to estimate the projected frontal surface area of complex three‐dimensional shapes such as flexible freshwater macrophytes at different flow conditions
Author(s) -
Sagnes Pierre
Publication year - 2010
Publication title -
limnology and oceanography: methods
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.898
H-Index - 72
ISSN - 1541-5856
DOI - 10.4319/lom.2010.8.474
Subject(s) - scale (ratio) , macrophyte , flow (mathematics) , drag coefficient , statistics , surface (topology) , software , mathematics , environmental science , hydrology (agriculture) , drag , computer science , cartography , geometry , geology , ecology , engineering , geography , biology , geotechnical engineering , aerospace engineering , programming language
For different purposes, the projected frontal surface area (A f ) of various organisms is often measured through image analysis. Such A f estimates are particularly used to assess the drag coefficient (C d ) in hydro‐ or aerodynamic studies of organisms. So far, estimates of A f from image analysis were generally biased (as subsequent C d assessments) for flexible organisms at a given flow, as these estimates were obtained using the same scale for parts of the body being close to or far from the camera. To assess this problem, I used image analyses on 3 flexible freshwater macrophytes, each exposed to 3 different flows. For these, using a single image scale resulted in ~20% errors of A f estimates if compared with a multiple scale (16 subscales, providing the “reference area”) assessment. Using 4 subscales and a human‐controlled plant contour definition for A f estimates was a good trade‐off between accuracy of A f estimates (error < 3% if compared with the reference area) and analysis effort (6 to 10 min for 1 plant in 1 flow condition). In comparison, using an automated definition of the plant contour (through computer software) and a composite scale (i.e., the mean of the 4 previous subscales) provided slightly worse A f estimates (error < 5% if compared with the reference area). Using the automatic procedure reduced the time to estimate A f for 1 plant in 1 flow condition by ~30%. Therefore, future studies requiring accurate A f assessments of flexible, complex, and elongated bodies should use an appropriate number of subscales as reference.