cence of chlorophyll 
irescence is optically 
means of an Optical 
ictively. 
i low divergence (0.4 
J. The choice of the 
is a high chlorophyll 
fe wl operation 
imer laser, allows to 
i, frequency shifting 
, due to the eye-safe 
m from the receiving 
sight. The laser spot 
shed by slowly rotat- 
de the glasshouse of 
n 3 ) under fully con- 
xl m 3 ) allowing for 
e glasshouse. 
> air mixed with 120 
pence samples group, 
dal chamber into the 
ed to 0 3 for a period 
d the last unpolluted 
responding reference 
tied from Pisa to our 
osure times for each 
Group 
I 
n 
m 
IV 
v 
VI 
vn 
vm 
"dT 
"IT 
In order to properly compare spectra arising from different measurements, normalization of the 
raw data acquired by the OMA have been performed. The spectra summed over a number of 100 or 200 laser 
shots have been normalized for an equivalent number of laser shots (100), and correction for the spectral re 
sponse of the photosensitive array detector has been taken into account Due to the high reproducibility of the 
pulse temporal shape and energy emitted by our laser, different averaged spectra are assumed to be normalized 
with respect to the laser energy without any further processing. 
The features of the spectra considered and analyzed in the present work are the F690/F740 spec 
tral ratio, the F450/F690 ratio and the total blue and total red fluorescence, which carry most of the information 
about the plant physiological status and photosyntetic activity . The content of each spectral band, needed for 
red/red and blue/red ratios, results from integrating the LEF spectrum with a 20 nm bandwidth around the band 
centre wavelength. 
3 - RESULTS & DISCUSSION 
3.1. General results 
Some interesting observations arise from the analysis of the experimental results. In figs, la-b we report the T IF 
spectra (excitation @ 308 nm), of Lolium plants exposed to S0 2 and to 0 3 , respectively, together with the cor 
responding control samples spectra. Variations due to gaseous pollution are found both in the red/red ratio and 
in the total blue and red fluorescence intensity. The blue fluorescence provides additional informations in the 
definition of stress sensitive regions in the space of spectral ratios. 
Fig. 1 LIF spectra of Lolium plants (control: solid line; polluted: dashed line). Excitation @ 308 nm. a) exposed 
to 120 ppb of S0 2 ; b) exposed to 100 ppb of 0 3 
By comparison, LIF spectra resulting from excitation @ 421 nm exhibit only the chlorophyll-a 
(Chl-a) red fluorescence , while the blue fluorescence around 450 nm is almost absent (fig. 2). 
After normalization over the laser pulse energy and spot area, die excitation efficiency vs wl have 
been measured for different herbaceous species, including also allium and malva. Data for lolium and spinach 
show that the red fluorescence following the 421 nm excitation is much larger (about a factor 60) than the corre 
sponding emission at 308 nm excitation 
By comparing the corresponding LIF spectra and spectral ratios, it turns out that 0 3 is by far more 
effective than S0 2 as plant pollutant, also taking into account the shorter exposition time and the lower concen 
trations used for the former gas. This fact may be due to the decreased concentration of C0 2 in an ozone pol 
luted atmosphere. 
Another important feature, already noticed in field campaigns [4] and confirmed in this experi 
ment, is that in general a plant stress status is reflected both in an increased red/red spectral ratio and in an in