====== Setup for SCORE-MIP radiative transfer model intercomparison ====== {{:intercomparisons:scoremip:task1a_rt_comparison_march2010.pdf|Download pdf}} [[:intercomparisons:scoremip:results|See results]] ==== Geophysical scenarios ==== * for 5 different aerosol scenarios: - Low aerosol loading (AOD ~ 0.05); no cirrus cloud: use gas + Rayleigh + aerosol (low_aod) - Low aerosol loading (AOD ~ 0.05) with cirrus cloud: use gas + Rayleigh + aerosol (low_aod) + cirrus - High aerosol loading (AOD ~ 0.3) no cirrus cloud: use gas + Rayleigh + aerosol (high_aod) - High aerosol loading (AOD ~ 0.3) with cirrus cloud: use gas + Rayleigh + aerosol (high_aod) + cirrus - No aerosol (pure Rayleigh): use gas + Rayleigh * Geometry - Solar zenith angle: SZA = 10, **45**, 70 - Viewing angle: nadir and sunglint * Surface - Lambert albedo (same in each band): 0.05, **0.1**, 0.3 - Ocean sunglint: windspeed 5 m/s with viewing zenith angle = solar zenith angle ==== Setup of input layer data ==== * For each layer of a 60 layer atmosphere: - Optical depth of gaseous absorption, aerosol, cirrus and Rayleigh extinction - Single scattering albedo of aerosols and cirrus - Scattering matrix for aerosols, cirrus and Rayleigh (layer independent) * Spectral range: - Band 1: O2 A-band: - **Window w1: 13080 – 13090 cm-1** (1000 spectral points) - Window w2: 12976 – 12980 cm-1 (400 spectral points) - Band 2: 1.61 micron CO2: - Window w1: 6202 – 6206 cm-1 (400 spectral points) - Band 3: 2.04 micron CO2: - Window w1: 4841-4845 cm-1 (400 spectral points) - **Window w2: 4815–4819 cm-1** (400 spectral points) ==== RT setup ==== * The RT calculations for each scenario will be carried out for 4 setups of the RT code: - 'scalar accurate': scalar RT with standardized setups - 'scalar fast': fast version of scalar RT code - 'vector accurate': vector RT with standardized setups - 'vector fast': fast version of vector RT code * Parameters for ‘accurate’ setup: 32 full streams, delta-M on, single scattering correction on, plane parallel, 60 vertical layers * ‘Fast’ setup: here the code should be run in a configuration so that the RT calculations are sufficiently fast for the retrieval of a large number of spectra * Groups can also provide the RT calculations only for some of the 4 setups. ==== Reduced set of test cases ==== All combinations of parameters result in a very large set of spectra to be compared. As a first step we evaluate only SZA=45°, albedo=0.1 or BPDF, band 1 (window 1) and band 3 (window 2). This results in the following 20 test cases: | | b1 w1 Lambert| b1 w1 Ocean | b3 w2 Lambert | b3 w2 Ocean| |low aerosol | 1a | 1b | 1c | 1d | |low aerosol + cirrus | 2a | 2b | 2c | 2d | |high aerosol | 3a | 3b | 3c | 3d | |high aerosol + cirrus | 4a | 4b | 4c | 4d | |pure Rayleigh | 5a | 5b | 5c | 5d | ==== Input data ==== Please download the input data: {{:intercomparisons:scoremip:data.zip}}. * The data directory contains subdirectories for aerosol, cirrus, gas and Rayleigh that contain files with the optical depth, single scattering abedo and scatteringmatrix * Optical depth for is given for each spectral point and each vertical layer. The uppermost layer and the last column is the optical depth of the bottom layer. * Single scattering albedo (ssa) for aerosol and cirrus is given for each spectral point * The input files are given for each spectral band separately. They will be identified with b1, b2, b3 for the O2 A Band, the 1.6 micron Band and the 2.03 micron band followed by the window number (w1 or w2). * We assume that the scattering matrix is constant across each band, ie. only one scattering matrix is given per band. The scatteringmatrix file contains all 3 bands * Scattering matrix is given as alpha1-4,beta1-2 according to following representation (Note that alpha1 is the scalar phase function): * Rayleigh depolarization factor 0.0279 * Please note that the forward peak of cirrus phase function has been truncated and the other optical properties of cirrus (single scattering albedo) have been modified accordingly. ==== Organization of output data ==== * Give output as 2 column ASCII files with wavenumber and I or I-Q * Generate separate file for each geophysical scenario * Include header that provides information about the run (RT setup, code version etc). * Start header lines with # * Provide the computation time for each scenario together with the computation time for the test programme (tespol.f90)