Use of hyperspectral reflectance measurements to predict nitrogen sufficiency of cereal and potato crops using a physical based approach

Optimizing nitrogen (N) fertilization by in-season measurements of crop N status may improve fertilizer N-use efficiency. This study evaluated the feasibility of using the inverted PROSAIL (PROSPECT+SAIL) reflectance model to predict leaf N status from canopy reflectance hyperspectral data acquired over spring wheat and potatoes.;The inverted PROSPECT model could not predict protein content with any degree of accuracy for either of the crop species. Leaf chlorophyll content was predicted with reasonable accuracy throughout the season for potato, particularly in 2002, but only at Zadoks Growth Stage (ZGS) 35 and later for wheat. Leaf chlorophyll and protein content for both crops correlated well on individual sampling dates.;For both crops, the inverted PROSAIL model was able to estimate leaf chlorophyll content with increasing accuracy over time until complete canopy closure occurred. For potatoes, the estimation accuracy declined when the potato vines subsequently collapsed, exposing more soil background. Similar soil background interferences occur at early growth stages for wheat. In all cases, leaf chlorophyll content was systematically over-estimated and LAI was systematically under-estimated, because of an inter-correlation effect within the model between these two parameters. The synthetic canopy chlorophyll content, which is the product of leaf chlorophyll content with LAI, was estimated with better accuracy than each of these variables individually.;The good leaf chlorophyll content estimation when the canopy is closed and LAI values are below 4.5, and the good relationships between leaf chlorophyll and N contents, suggest that the model may be used to predict potato and wheat N status.;Trials were conducted on two potato cultivars under seven different N fertility treatments (0 - 300 kg N ha-1) in 2001 and 2002 and one wheat variety under six different N fertility rates (0 – 200 kg N ha-1) in 2002. Leaf reflectance and transmittance, canopy reflectance, leaf chlorophyll content, leaf protein content and leaf area index (LAI) were measured.