Minimizing errors in LAT estimates from laser-probe inclined-point quadrants

Leaf area index (LAT) measurement by the method of inclined point quadrants has recently been automated using laser-induced chlorophyll fluorescence to detect green leaves. The simplest implementation of this approach assumes (i) leaves are randomly distributed, both in azimuth and in horizontal position, and (ii) the laser beam is infinitely thin. These assumptions are violated by heliotropic plants, crops grown in rows, and real world lasers. Mathematical modeling was used to analyze the resulting errors and to evaluate possible solutions. One model treated crop rows as rectangular prisms filled with leaves categorized by inclination and azimuth angle. Nonrandom leaf azimuth and crop arrangement in rows were predicted to decrease the accuracy of laser-probe LAT estimates. It was also shown that such errors can be reduced to < 15% by averaging LAT estimates from an appropriate range of laser beam azimuths. Sample size was found to be most critical for LAT < 0.5 or > 4. A second model predicted the laser-induced chlorophyll florescence signal (assumed proportional to laser-illuminated leaf area) as a circular laser spot was scanned across a rectangular leaf. Laser beam diameter was predicted to be a significant source of error whenever the radius of the laser beam exceeded the leaf width. This prediction was confirmed experimentally, using rice seedlings growing underwater.