Multispectral retrieval (satellite)

Minimal input

Input is provided in ./src/Input_data.xlsx that has several sheets.

To run the model for multispectral instruments use main_sat.m.

sheet	purpose	action
Input	Parameters and ranges	Set 1 in the column tune in front of the parameter you want to retrieve Change value of fixed parameters (where tune == 0)
Satellite	Multispectral retrieval	Provide image_path (NetCDF4) Select instrument_name Provide band names used in your NetCDF for angles and coordinates Go to Bands sheet
Bands	Band names and wl	Provide your_names of bands as they are in NetCDF Provide your_wl of bands (for plotting) Keep the order! => leave a gap in bands if you excluded any

Retrieval on image subset (K x K)

If your image is big but you are interested only in a small area around a certain point you can specify the coordinates of that point and the number of pixels around it.

sheet	purpose	action
Satellite	Multispectral retrieval	Provide pix_lat, pix_lon Provide K (number of pixels around that point)

Parallel computing (parfor)

Fully supported as for hyperspectral retrieval Parallel computing (parfor).

The changes should be made in main_sat.m (currently lines 105-111)

Output

Windows

If you run main_sat.m you provide input file (image_path) as NetCDF4, the output will have two files:

1. %Y-%m-%d_%H%M%S.xlsx (see hyperspectral output Windows)

2. NetCDF4: retrieved parameter (Cab, LAI etc.) are be written as separate bands

So you can have a map of your retrieval.

Linux

Nobody forbids you to do the same trick that is used for hyperspectral retrieval, just copy these lines from main.m to main_sat.m

main.m

%% start saving
q = parallel.pool.DataQueue;
if isunix
    path = io.create_output_folder(input_path, path, tab.variable);

    tmp_zeros_res.rmse = rmse_all;
    tmp_zeros_res.parameters = parameters;
    tmp_zeros_unc.std_params = parameters_std;

    tmp_zeros_res.refl_mod = refl_mod;
    tmp_zeros_res.soil_mod = refl_soil;
    tmp_zeros_res.sif = sif_rad;
    tmp_zeros_res.sif_norm = sif_norm;

    tmp_zeros_meas.refl = refl_mod;
    tmp_zeros_meas.wl = measured.wl;
    tmp_zeros_meas.i_sif = measured.i_sif;

    io.save_output_csv(0, tmp_zeros_res, tmp_zeros_unc, tmp_zeros_meas, path)
    afterEach(q, @(x) io.save_output_csv(x{1}, x{2}, x{3}, x{4}, path));
else
    path = io.create_output_file(input_path, path, measured, tab.variable);
    afterEach(q, @(x) io.save_output_j(x{1}, x{2}, x{3}, x{4}, path));
end

Multispectral sensor from .txt file

If you do not want to retrieve from NetCDF4 file you can provide reflectance as .txt file and a proper sensor name.

Input is provided in ./src/Input_data.xlsx.

To run the model for multispectral instruments from .txt use main.m.

sheet	purpose	action
Input	Parameters and ranges	Set 1 in the column tune in front of the parameter you want to retrieve Change value of fixed parameters (where tune == 0)
Filenames	Multispectral retrieval	Provide path to measured reflectance (text) Provide path to wavelength of measurements reflectance_wl Select instrument_name Chose the number of the spectrum you fit c or all (-999) Provide geometry tts, tto, psi or lat, lon, datetime, tz Go to Bands sheet
Bands	Band names and wl	Provide your_names of bands Provide your_wl of bands (same as reflectance_wl) Keep the order! => leave a gap in bands if you excluded any