Impacts of coverage and canopy water depth on the spectral characteristics for a submerged plant Cabomba caroliniana

Remote sensing technology can be used to timely monitor the distribution and growth status of submerged aquatic vegetation(SAV) on a large scale.However,the water depth from water surface to plant canopy may affect the spectral characteristics of submersed plants,resulting in an inaccuracy to interpret the related remote sensing images.In this study,the spectral characteristics of a submerged plant Cabomba caroliniana with varying coverage and canopy water depth were measured using a FieldSpecTM Pro JR Spectroradiometer in the field and a control experiment.The influence of canopy water depth on the spectral reflectance of Cabomba caroliniana was analyzed and a model was established to deduce the coverage of Cabomba caroliniana community based on the spectral reflectance and the canopy water depth.The results showed that(1) the basic characteristics of spectral reflectance for varying coverage of Cabomba caroliniana appeared mainly at the bands of green and NIR;(2)the spectral reflectance of Cabomba caroliniana community showed a negative correlation with the canopy water depth,and the spectral reflectance of Cabomba caroliniana community with same coverage decreased with increasing canopy water depth,especially at the band of NIR;(3)the shallower the canopy water depth,the more significant the correlation between the coverage and the spectral reflectance of Cabomba caroliniana community,which enhanced with the increasing coverage of Cabomba caroliniana;(4) the most significant correlations between the coverage and the spectral reflectance of Cabomba caroliniana community were observed at the band of 692—898 nm,while the most significant correlations between the canopy water depth and the spectral reflectance of Cabomba caroliniana community were near the bands of 710 nm and 806 nm;(5) a revised model established at the bands of 710 nm and 806 nm and adopted both the parameters of canopy water depth and spectral reflectance was much better than a simple model established at the bands of 710 nm and 806 nm and adopted only the parameter of spectral reflectance to deduce the coverage of Cabomba caroliniana community.The determination coefficients(R2) between the predicted and measured coverage of Cabomba caroliniana community were 0.742 and 0.777 for the simple model at the band of 710 nm and 806 nm,while were 0.852 and 0.935 for the revised model at the band of 710 nm and 806 nm.The revised model adopted both the parameters of canopy water depth and spectral reflectance could effectively reduce the influence of canopy water depth on the deductive accuracy for the coverage of submerge vegetation.The results from this study could be helpful to timely and accurately monitor the distribution and dynamics of submerged vegetation on a large scale using remote sensing technology.