Spectral estimates of net radiation and soil heat flux

Abstract

Conventional methods of measuring surface energy balance are point measurements and represent only a small area. Remote sensing offers a potential means of measuring outgoing fluxes over large areas at the spatial resolution of the sensor. The objective of this study was to estimate net radiation (R n) and soil heat flux (G) using remotely sensed multispectral data acquired from an aircraft over large agricultural fields. Ground-based instruments measured R n and G at nine locations along the flight lines. Incoming fluxes were also measured by ground-based instruments. Outgoing fluxes were estimated using remotely sensed data. Remote R n, estimated as the algebraic sum of incoming and outgoing fluxes, slightly underestimated R n measured by the ground-based net radiometers. The mean absolute errors for remote R n minus measured R n were less than 7%. Remote G, estimated as a function of a spectral vegetation index and remote R n, slightly overestimated measured G; however, the mean absolute error for remote G was 13%. Some of the differences between measured and remote values of R n and G are associated with differences in instrument designs and measurement techniques. The root mean square error for available energy (R n - G) was 12%. Thus, methods using both ground-based and remotely sensed data can provide reliable estimates of the available energy which can be partitioned into sensible and latent heat under nonadvective conditions.