Irradiance calibration data supplied by Ocean Optics is expressed in energy units per detector count. In other words needs to be corrected for the wavelength step between detector pixels and for the area of the cosine diffuser used. The equation to use is \[E_\lambda = \frac{k_\lambda \cdot c_\lambda}{A \cdot t \cdot \Delta \lambda}\] where \(k_\lambda\) is the calibration constant for a detector pixel, and \(c_\lambda\) are the counts registered by the corresponding detector pixel, \(A\) is the area of the diffuser, \(t\) is the integration time and \(\Delta \lambda\) is the wavelength step for the pixel.

With \(k_\lambda\) in \(\mu J c^{-1}\), \(A\) in \(mm^2\), \(t\) in seconds and \(\Delta \lambda\) in \(nm\), the result as spectral energy irradiance in \(W m^{-2} nm{-1}\)

We can substitute \(c_\lambda / t\) by “counts per second”, which is what we use in the computations, and recalculate the remaining of the equation as a new set of calibration constants.

\[k_\lambda^\prime = \frac{k_\lambda}{A \cdot \Delta \lambda}\]