Descriptor File Structure¶
This file contains microphysical specifications for all hydrometeors. It includes central parameters for PSD, m-D, A-D relation and selection of scattering model for frozen hydrometeors. All values (e.g. a and b parameter for m-D relation) have to be provided in SI units.
The descriptor file consists of the following fields
| field | values | description |
|---|---|---|
| name | string | name of the hydrometeor |
| as_ratio | positive float | aspect ratio of hydrometeor. <0 means oblate. Note that mie-sphere, liudb and rayleigh do not consider as_ratio when estimating the scattering properties. |
| liq_ice | -1, 1 | Phase of hydrometeor, -1 is ice, 1 is liquid |
| rho_ms | positive float | Density of hydrometeor (ice only) |
| a_ms | positive float | ‘a’ parameter of mass-size relation (ice only) |
| b_ms | positive float | ‘b’ parameter of mass-size relation (ice only) |
| alpha | positive float | ‘alpha’ parameter of cross section area-size relation (ice only). Only required for the advanced radar simulator. |
| beta | positive float | ‘beta’ parameter of cross section area-size relation (ice only). Only required for the advanced radar simulator. |
| moment_in | 0,1,2,3,12,23,13 | Moment provided in input file (see below) |
| nbin | positive int | Number of discrete size bins (internally, always nbin+1 is used) |
| dist_name | str | Name of hydrometeor distribution (see table below) |
| p_1 | float | 1st parameter of hydrometeor distribution (see table below) |
| p_2 | float | 2nd parameter of hydrometeor distribution (see table below) |
| p_3 | float | 3rd parameter of hydrometeor distribution (see table below) |
| p_4 | float | 4th parameter of hydrometeor distribution (see table below) |
| d_1 | positive float | minimum diameter |
| d_2 | positive float | maximum diameter of size distribution (not required for mono-disperse distribution) |
| scat_name | string | Scattering model. For passive and active, it can be “disabled” for disabling the specific hydrometeor, “mie-sphere”, “tmatrix” (very slow), “liudb%P”, and “hongdb%P”. For active only: “rayleigh”, “rayleigh-gans”, “ss-rayleigh-gans_%k_%b”. %P is for the particle type of the Liu or the Hong database (00 to 10) as described in their publications (see PAMTRA publication), %k and %b are for the kappa and beta parameter of the self similar Rayleigh Gans approach. If not provided, default values kappa = 0.19 and beta = 0.23 are used. |
| vel_size_mod | string | Used model to estimate the fall velocity of hydrometeors. Can be ‘khvorostyanov01_drops’ (recommended for liquid), ‘khvorostyanov01_spheres’, ‘rogers_drops’, ‘heymsfield10_particles’ (recomended for ice and snow), ‘khvorostyanov01_particles’, ‘rogers_graupel’, ‘powerLaw_%A_%B’, ‘corPowerLaw_%A_%B’ (with pressure correction). %A and %B are teh parameters of the power law v(D) = %A*D^%B. Only required for the advanced radar simulator. |
| canting | float | Canting angle of hydrometeors. Only for Tmatrix and (Self similar) Rayleigh Gans. The latter accept only values of 0 and 90deg. |
The complete list of options which can be used to define the PSD is given in table below. The table should be used as follows: given a distribution dist_name and a set of the parameters p_1 to p_4, moment_in describes the available choices for the moments of the PSD that have to be included in the input_file. 1 means the total number concentration, 2 the effective radius, 3 the mass concentration, 13 both number and mass concentration and so on. d_1 and d_2 describe the diameter of the smallest and largest particle considered to calculate absorption and scattering properties for a given class of hydrometeors.
| dist_name | p_1 | p_2 | p_3 | p_4 | moment_in | d_1 | d_2 |
|---|---|---|---|---|---|---|---|
| mono | -99 | -99 | -99 | -99 | 1 or 3 | d_1 | -99 |
| mono | N_T | -99 | -99 | -99 | 2 or 3 | -99 | -99 |
| mono_cosmo | -99 | -99 | -99 | -99 | 3 | -99 | -99 |
| exp | N_T | -99 | -99 | -99 | 2 or 3 | d_1 | d_2 |
| exp | -99 | R_eff | -99 | -99 | 1 or 3 | d_1 | d_2 |
| exp | -99 | -99 | N_0 | -99 | 1, 2, or 3 | d_1 | d_2 |
| exp | -99 | -99 | -99 | -99 | 12, 13, or 23 | d_1 | d_2 |
| exp_field_t | -99 | -99 | -99 | -99 | 2 or 3 | d_1 | d_2 |
| exp_field_tq | -99 | -99 | -99 | -99 | 3 | d_1 | d_2 |
| exp_ryan | -99 | -99 | -99 | -99 | 1 or 3 | d_1 | d_2 |
| logn | N_T | -99 | sigma | -99 | 2 or 3 | d_1 | d_2 |
| logn | N_T | R_eff | -99 | -99 | 3 | d_1 | d_2 |
| logn | -99 | R_eff | sigma | -99 | 1 or 3 | d_1 | d_2 |
| logn | -99 | -99 | sigma | -99 | 12, 13, or 23 | d_1 | d_2 |
| mgamma | N_T | -99 | mu | gamma | 2 or 3 | d_1 | d_2 |
| mgamma | -99 | R_eff | mu | gamma | 1 or 3 | d_1 | d_2 |
| mgamma | -99 | -99 | mu | gamma | 12, 13, or 23 | d_1 | d_2 |
Note that it is also possible to provide discrete vectors for particle size concentration, mass, cross section area, density, and aspect ratio. See
addFullSpectra of the class pamDescriptorFile.
An exemplary descriptor_file for a simulation with two hydrometeors is given in the table below. For the first hydrometeor category named cwc_q, the log-normal distribution have been used with a sigma of 0.38, fixed for the whole simulation via the parameter p_3. Since the hydrometeor is in liquid phase, the density and the a and b parameters of the mass-size relation will be ignored. The aspect ratio and canting angle has been set to missing value (-99) because single scattering properties are calculated with Mie theory assuming spherical particles. The second hydrometeor swc_q is used to simulate snow. In this case an exponential distribution is used with a temperature dependent intercept parameter. The thermodynamic state of the atmospheric columns included in the simulation and the surface properties are provided via the input_file (IF), which can be in ascii or net-cdf format. The minimum set of parameters that has to be included for clear sky simulations are: surface temperature, longitude, latitude, date and profiles of temperature, pressure, height and relative humidity. In case of cloudy profiles, the moments of the PSD specified in the descriptor_file need to be included. The PAMTRA model gives as output TB for 2 observational heights, one is fixed at ground and the other can be specified in the IF for each profile.
| name | as_ratio | liq_ice | rho_ms | a_ms | b_ms | alpha | beta | moment_in | nbin | dist_name | p_1 | p_2 | p_3 | p_4 | d_1 | d_2 | scat_name | vel_size_mod | canting |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| cwc_q | -99. | 1 | -99. | -99. | -99. | -99. | -99. | 23 | 100 | logn | -99. | -99. | 0.38 | -99. | 1.e-12 | 1 .e-2 | mie-sphere | khvorostyanov01_drops | -99. |
| swc_q | -99. | -1 | -99. | 0.038 | 2.0 | 0.3971 | 1.88 | 3 | 100 | exp_field_t | -99. | -99. | -99. | -99. | 0.51e-10 | 2 .e-2 | mie-sphere | heymsfield10_particles | -99. |