Estimation of amount of volcanic ashes piled on plant leaves by spectral reflectance

Abstract

A non-contact and non-destructive method was applied to the estimation of volcanic ashes on the surface of soybean plant leaves using spectroradiometer.
In this paper, we found the most effective spectral band and experimental equations suitable for estimating amount of volcanic ashes (white ashes and black ashes from Mt. Sakurajima Volcano) piled on soybean plant leaves.
The most effective spectral index for estimting the white volcanic ashes piled on plant leaves (Y_W, g/m₂) was the value of (R_900nm-R_450nm)/(R_900nm+R_450nm). The regression equation was Y_W=244.Oe^(-4.60x)(X:dimensonless). The determination coefficient (r²) was 0.939.
The most effective spectral index for estimating the black volcanic ashes piled on plant leaves (Y_B) was the value of (R_800nm-R_450nm)/(R_800nm+R_450nm). The regression equation was Y_B=252.8e^(-4.22X) (X:dimensionless). The determination coefficient was 0.965.
The observed and the estimated values of the amount of volcanic ashes (white ashes and black ashes) piled on plant leaves were well distributed on the ratio line of one to one. The determination coefficients were 0.958 and 0.970, and the standard errors of prediction were 4.78 and 4.13 respectively.