= Satellite test of greenhouse spectral change
A <satellite test of greenhouse spectral change> compares outgoing <radiance> with <radiative transfer> predictions. The 2001 study compared IRIS on Nimbus 4 (1970) and IMG on ADEOS (1997). Matching April–June ocean regions, masking land, screening clouds and correcting resolution and <wavenumber> controlled comparability. Spectral <radiance> became <brightness temperature> readouts. Regional comparisons and independent calculations tested interpretation. <Carbon dioxide> and <methane> features supported changed absorption; residual clouds, calibration and interannual variability limited background inference.<Increases in greenhouse forcing inferred from the outgoing longwave radiation spectra of the Earth in 1970 and 1997 by John E. Harries and others>[Harries et al., 2001, pp. 355–357, Table 1, Fig. 1 and Methods]{ref}
This observational comparison tests <radiative transfer>; it does not by itself measure equilibrium <climate sensitivity> or validate a complete <climate model>.
To understand the readout, write an instrument's spectral channel schematically as
$$
y_j=\int K_j(\widetilde\nu)I(\widetilde\nu,\widehat{\mathbf n})\,d\widetilde\nu+b_j+e_j,
\qquad \int K_j(\widetilde\nu)\,d\widetilde\nu=1.
$$
Here $I$ is spectral <radiance> in the viewing direction $\widehat{\mathbf n}$, $K_j$ is a normalized channel response, $b_j$ is calibration offset and $e_j$ is random error. Subtracting two channel readings isolates an atmospheric difference only after their response functions and sampling are comparable. Matching kernels reduces a false difference caused merely by observing a narrow <spectral line> at different resolutions; it cannot remove every calibration offset.
<Outgoing longwave radiation> in an <energy> budget also integrates <radiance> over upward directions and over the thermal spectrum, with the appropriate projected-angle factor. A <brightness temperature> converts one spectral <radiance> to the equivalent <blackbody> value through <Planck's law>; it is not automatically surface <temperature>. This explains why a spectral difference, a broadband <radiative flux> difference and surface warming are related observables with different units and different inferences.
Back to article page
