Estimation of fuel moisture content by inversion of radiative transfer models to simulate equivalent water thickness and dry matter content:: Analysis at leaf and canopy level

Fire danger models identify fuel moisture content (FMC) of live vegetation as a critical variable, since it affects fire ignition and propagation. FMC can be calculated by dividing equivalent water thickness (EWT) by dry matter content (DM). The leaf optical properties spectra (PROSPECT) model was inverted to estimate EWT and DM separately using the Leaf Optical Properties Experiment (LOPEX) database, based on 490 measurements of leaf optical and biochemical properties. DM estimations were poor when leaf samples were fresh (r(2) = 0.38). Results of a sensitivity analysis conducted on the spectral response of samples demonstrated that water absorption masks the effects of DM on the spectral response. This causes a poor estimation of FMC (r(2) = 0.33), even though EWT estimation was good (r(2) = 0.94). However, DM of dry samples was accurately estimated (r(2) = 0.84), since no water was present. FMC estimation in fresh leaf material improved considerably (r(2) = 0.89) by accounting for DM in fresh samples (r(2) = 0.71), when a constant species-dependent DM value is used. A similar approach was taken on a canopy level by linking the PROSPECT leaf model with the Lillesaeter infinitive reflectance canopy model using data from laboratory measurements under controlled conditions. As expected, results indicate grater difficulty to estimate DM and FMC on a canopy level, although similar trends were observed. DM estimation improved from r(2) = 0.12 to r(2) = 0.39 when considering measurements of dry samples, which when used for FMC estimation, correlations increase from r(2) = 0.62 to r(2) = 0.81. Therefore, DM can be accurately estimated only when plant material is dry, and it is a necessary measurement in order to estimate FMC accurately. DM remains stable over annual time periods although lower values are expected during the drought season. However, time variations in DM are smaller than DM variations among species. Because there is also a decrease in water content with low EWT values during phenologically dormant periods, multitemporal data could be used to estimate FMC. Further research must be applied on real canopies to confirm these results.