creased blue fluorescence, although this is not true for the group of plants treated twice with 0 3 . This seem to 
indicate that the validity of F690/F740 ratio as a stress indicator is limited to only very little stresses. When 
highly polluted samples are examined, more complex physiological phenomena involving leaf structure perma 
nent damages may be involved. 
Fig. 2 LIF spectra of Lolium plants (control solid line and polluted dashed line). Excitation @421 nm. The 
plants are exposed to 100 ppb of 0 3 . 
3.2. Fluorescence response vs wavelength 
By taking advantage of the Raman shifter output beams, we were able to excite the chlorophyll fluorescence at 
different wavelengths. Main results for allium and Malva are reported in figs. 3 and 4. As far as the red fluores 
cence intensity is concerned, it is clearly seen to increase with increasing wavelength (fig. 3). Such an effect is 
probably due not only to an increasing chlorophyll pigments absorption cross section, which is a maximun 
around 450 nm, but also to a larger penetration depth of the laser light into the illuminated leaves, so that a 
greater number of chlorophyll molecules can fluoresce. If the red fluorescence vs wl seems to be easily under 
stood, explanation of the Rj/R 2 ratio behaviour, shown in fig. 4 for allium and Malva is more difficult. Actually 
and at least for the samples considered, the dependence of the R,/R 2 ratio vs wl seems typical for each investi 
gated herbaceous species. Therefore it is of great importance for the analysis of fluorescence experimental data 
is the exciting laser wl which strongly affects the Rj/R 2 ratio, so that it is mandatory to associate to each deter 
mination of R,/R 2 ratio the excitation wl before attempting any comparison with different lidar systems. 
33. LIF spectra vs laser energy 
Another important parameter to take into account in designing an efficient lidar fluorosensor is laser the pulse 
energy, since the LIF SNR increase may saturate with increasing pulse energy and signal distortion may appear 
due to the occurrence of annealing effects. A complete analysis of this topic is not performed in the present 
work, where we were interested only to operate out of the saturation region. In order to check this point, exten 
sive measurements on different herbaceous species vs the laser pulse energy have been carried out, and typical 
results obtained for basil are reported in fig. 5. Observation of these data seems to indicate that for all the energy 
range used there is no signal saturation.