Multisite analyses of spectral-biophysical data for corn
1990
Wiegand, C.L. | Gerbermann, A.H. | Gallo, K.P. | Blad, B.L. | Dusek, D.
Reflectance factors and biophysical plant measurements for corn (Zea mays L.) experiments conducted at Bushland, Texas (102.2°W, 35.2°N), Tryon, Nebraska (100.8°W, 41.6°N), W. Lafayette, Indiana (87.0°W, 40.5°N) and Weslaco, Texas (98.0°W, 26.2°N) were fit by various equation forms for four of the most used vegetation indices (VI): n-space greenness (GVI), normalized difference (NDVI), perpendicular (PVI), and near-infrared to red ratio (RVI). The objective was to produce relations from the data pooled across all locations that could be recommended for general use for corn. Data were analyzed by premaximum leaf area (pre-Lmax), post-maximum leaf area (post-Lmax), and whole season portions of the growing season. The vegetation indices ranked in descending order: GVI, PVI, NDVI, and RVI based on coefficients of determination (R2), root mean square error (RMSE) in estimating leaf area index (L), and robustness across locations, soils, sun angles, cultivars, and radiometers. The power form was as good as any other equation form for the two-band vegetation indices, but quadratic and exponential equation forms were better for the three- or four-band GVI. Fractional absorbed photosynthetically active radiation (FPAR) was described for the combined data from Weslaco (PVI) and West Lafayette (GVI) by FPAR = 0.0088 + 0.0315 (PVI, GVI) (R2 = 0.94, RMSE = 0.07) and FPAR = 1 - exp[0.400(L / cos Z)] (r2 = 0.95, RMSE = 0.04), where Z is sun zenith angle. We conclude that the corn canopies characterized by green leaf area index so dominated the reflectance factor, hence VI observations, that it was possible to develop general relations for estimating L and FPAR for corn from VI measured in widely separated experiments.
显示更多 [+] 显示较少 [-]