Minimize Water Vapour Retrieval

This algorithm, applied to the OLCI signals, aims to retrieve the atmospheric water vapour content of clear sky pixels.
Two different algorithms are applied depending on the brightness of the underlying surface: land and water in high glint conditions or covered by ice, or water outside high glint. Both algorithms are based on a differential absorption method using two spectral bands close to each other (one within the absorption band and the other outside the absorption band) allowing measurement of absorption due to water vapour along the light path. Both algorithms are implemented using a neural network method.
Over water, pixels not contaminated by high glint or floating ice, have local averaging over a sizeable window centred over the pixel performed to improve the signal to noise ratio in the NIR. The neural network takes as input, observation geometry, water vapour transmittance (ratio of reflectance at 900 nm and at 885 nm) and horizontal wind modulus, representative of sea surface roughness.
Over brighter surfaces, i.e. land masses and water contaminated by high glint or ice, the neural network takes as input, observation geometry, water vapour transmittance (ratio of reflectance at 900 nm and at 885 nm), normalised radiances at 885 nm and 754 nm, molecular oxygen transmittance (ratio of reflectances at 761 and 754 nm) and the wavelength of the oxygen absorption channel for the considered pixel.

 

Functional Block Diagram of Water Vapour Retrieval (PP-WV)