Temporal and spatial relationships of canopy spectral measurements

Abstract

Ground-based, remotely sensed reflectance and temperature data were collected over differentially irrigated, developing cotton and sorghum canopies in order to investigate interrelations of these parameters; to monitor their temporal changes ; to understand their spatial structure ; and to estimate crop coefficient (KO from canopy reflectance. Spectral reflectance and derived vegetation indices showed ability to significantly discriminate among differential irrigation levels of sorghum canopies, starting the fourth week of growth. All vegetation indices increased as a result of crop development, with the perpendicular vegetation index (PVI ) demonstrating the greatest potential for assessing water stress conditions, whereas, soil indices behaved independently of crop development and water stress. Canopy temperature and derived water stress indices, on the other hand, were in high concordance and were able to detect crop water stress with variable degrees of sensitivity. Experimental variograms revealed that cotton reflectance and temperature were not spatially dependent when all water treatments were included. For the moisture stress treatments, only canopy temperature exhibited spatial dependence early in the period of stress. Sorghum canopy reflectance and temperature demonstrated some spatial structure; however, a drift was suspected due to regularity in the data spatial distribution. Normalized difference (ND), normalized perpendicular vegetation index (NPVI) and normalized green vegetation index (NGVI), for fifty days covering the period from planting to heading, were fitted with a complementary error function equation with minor adjustment. Both NPVI and NGVI displayed a 1:1 relation with interpolated tabular values of basal Kc, whereas ND deviated from the 1:1 relation for the period beyond 30 days after planting. The model was also found to be valid for estimating K(c) for moderately deficit irrigation conditions.