Stand parameter estimation of artificial evergreen conifer forests using airborne images: An evaluation of seasonal difference on accuracy and best wavelength

The relationship between the stand parameters (top layer height (H1) and volume/ha (Vha)) and digital number (DN) were evaluated for evergreen conifer stands using three airborne images with 4-m spatial resolution, which were taken in June 1995, September 1993, and October 1994 using the Compact Airborne Spectrographic Imager (CASI). Estimation accuracy of the stand parameters, their seasonal changes, and suitable wavelength were analyzed using correlation coefficients and a regression analysis. The minimum DN of stands, which showed the darkness of a canopy shadow, had a higher correlation with H1 than the average and maximum DN while the average DN had a higher correlation with Vha. The green channels gave the highest correlation coefficients with H1 and Vha, which exceeded — 0.9 for the September and October images. However, the red channels had a consistently high correlation with the stand parameters for the three images. The near infrared channels gave poor correlations with H1 and Vha for the June image. Spectral variations among trees may affect the relationship between DN and the stand parameters in the leaf maturation period in June. Consequently, the late growing season was better at giving consistent results for the stand parameter studies. There was a linear relationship between the measured and the estimated stand parameters for the validation plots especially for the H1 case of September with sufficient accuracy. Nadir viewing images, which had high spatial resolution and a wide dynamic range such as the CASI images, were necessary to estimate the stand parameters accurately.