Minimize IMT Neural Net

The Inverse Modelling Technique (IMT) uses an Inverse Radiative Transfer Model-Neural Network (IRTM-NN) to estimate from normalised water-leaving reflectance at OLCI bands:

b400 to b674, b709, θs, θv and Δϕ the log10 of

  • the absorption coefficient of humic acids (detrital material)
  • the absorption coefficient of fulvic acids (Gelbstoff material)
  • the absorption coefficient of algal pigment (Chlorophyll)
  • the scattering coefficient of total suspended matter
  • the scattering coefficient of white particles.

The IMT considers the complex nature of water-leaving reflectance and its parameterisation avoids any iterative procedures. The multiple non-linear regression method in this approach leads to a high reduction in computing time and is therefore fast enough for operational mass production of Level-2 products, but it requires a careful and elaborate determination of the multiple coefficients (training phase). The IRTM-NN procedure is already programmed and requires only data input. The data outputs from this algorithm are water-inherent optical properties:

  • total backscattering coefficient BBP443
  • total absorption coefficient ATOT443
  • phytoplankton absorption coefficient APH443
  • coloured Detrital and Dissolved Material absorption coefficient ADG443

Further used to derive the water constituents:

  • Chlorophyll (Chl2) concentration
  • Total Suspended Matter (TSM) concentration.

A quality indicator (represented by the flag in the following figure), which measures the average difference between the reflectances observed and those derived from the water's inherent optical properties, through a forward model neural network.

 

Functional Block Diagram of the IMT Neural Net