Fcal(t) 
bj.D (t-t¡) 0 F(t) , 
where ^ denotes the convolution product and bj are weighting factors. These factors are required to account for 
the differences in the fluorescence intensity produced by each elementary flash as a result of the different 
properties of Dex(t) and Fex(t). In this second step, the fit is done on bj, Fj and ij parameters by the Marquardt 
search algorithm. 
A first validation of this approach has been done with synthetic data created by computer simulation 
[30]. A second test consisted in measuring real data with our LIDAR system using an "artificial tree" in which 
the positions and the fluorescence lifetimes of the leaves were known [28]. The new deconvolution method has 
been further applied with success to fluorescence decay data obtained with the LIDAR system on a real canopy. 
Fig. 7 shows the signals obtained on a sorghum canopy at an equivalent distance of 50 m. The fluorescence 
signal was averaged over 16 shots. A mean lifetime of 0.32 ns was computed as expected considering the low 
light conditions (fluorescence level near Fo). 
7. RECENT DEVELOPMENTS IN BLUE FLUORESCENCE 
Blue-Green fluorescence Chlorophyll Fluorescence (red) 
(nm) 
FIG. 8. Emission spectrum of a spinach leaf excited at 290 nm 
In addition to red fluorescence, plants emit blue fluorescence (BF) when excited by UV light (Fig. 8). This 
signal has received recendy an increasing interest. It was shown that its emission spectrum depends on plant 
type and plant anatomy. For instance, emission from adaxial side of bifacial leaves differs substantially from 
abaxial side. Blue fluorescence was also shown to depend on environmental factors experienced by the plant, 
like water stress or nutrient concentration in soil. BF is a complex sum of emission by several fluorophores. The 
substances that are candidates for BF can be divided in two main classes: 
- aromatic compounds located in the vacuole and cell walls of the epidermis and that are related to the secondary 
plant metabolism and UV-protection. 
- cofactors of the metabolism, pyridine and flavine nucleotides, which are directly related to the redox state and 
fluxes of the plant cell. 
New insight has been obtain on the contribution of pyridine nucleotides to the BF emission in isolated 
chloroplasts and mesophyll (leaves without epidermis) by evidencing a variable, light induced BF which can be 
unambiguously attributed to NADPH, (See Fig. 9) [32]. In spite of these results, BF from NADPH seems 
hardly detectable, in whole leaves. Table 1 shows the relative contribution of the different parts of the leaf. 
From the analysis of these data it is concluded that, in intact leaves, the contribution of NADPH to BF, should 
be less than 3% of the whole BF emission. 
The problem of heterogeneity of BF has also been approached by time resolved spectroscopy in the 
sub-nanosecond domain, to differentiate more precisely the components in whole leaves [25]. More recently the 
same approach was used for a comparative analysis of leaves, mesophyll and isolated chloroplasts [31]. This 
work points out several lines of evidence indicating the presence of flavin nucleotides in a 3-4 ns kinetic 
component at any given level of organization of the leaf: chloroplasts, mesophyl or intact leaf. On the other 
hand, a 8-10 ns component shows all the characteristics reflecting the presence of nicotinamide nucleotides. 
However, other components are even more fluorescent in the blue. Decay-associated spectra and 
comparative analysis showed that the fluorescence of intact leaves was dominated by a fast (=.3 ns) and a 
medium (=1 ns) components [31]. They have a blue maximum, and are probably composed of fluorescent 
870