Descriptor file setting

[Davide88]

Good afternoon,
I am simulating a 89GHz radiometric channel looking upward at zenith.
In a first attempt, I simulated only clear sky conditions disabling all hydrometeors effects (hydro_off in the example picture). Then I tryied to introduce hydrometeors effects, in particular CWC, IWC, RWC, SWC which I retrieved from copernicus ERA5 reanalysis for a specific site (28x28km resolution roughly).
I suppose there is not a generic setting for all atmospheric conditions (it would be too easy), nevertheless I was wandering whether there is some documentation somewhere where to understand better how to set up properly the descriptor file for the cloudy/rainy days simulation.

I attach the brightness simulation for a random day in Payerne, the _off case (top picture) is with all hydrometeors disabled, while the _on case (bottom picture) is with the hydrometeors enabled. The red dots are the PAMTRA-simualated hourly Tb, while the blue curve is the signal from a 89GHz radiometer upward looking. I can see that qualitatively the _on simulation somehow follows the blue curve, despite an offset most probably due to descriptor file setting. Indeed, clear sky part of the plot (bottom picture, from 10AM on) shows a good agreement between simulations and measurements using the _off setting as descriptor file.

I am looking at the following page for informations:
https://pamtra.readthedocs.io/en/latest/descriptorFile.html

If you have any suggestion on how to set up the descriptor file for such case of cloudy/rainy cases, it would be very helpfull: I would like for the simulations (red dots) to be as close as possible to the real measurements (blue curve).

If there is any ERA5 reanalysis importer, it would be great as well.
For completeness, I list hereby the descriptor file setting for the two cases:

Hydro off descriptor file:
descriptorFile = np.array([
('cwc_q', -99.0, 1, -99.0, -99.0, -99.0, -99.0, -99.0, 3, 1, 'mono', -99.0, -99.0, 2.0, 1.0, 2.0e-6, 8.0e-5, 'disabled', 'khvorostyanov01_drops', -99.0),
('iwc_q', 1.0, -1, 700.0, -99.0, -99.0, -99.0, -99.0, 3, 1, 'mono', -99.0, -99.0, 1.564, 0.8547, 1.744e-5, 9.369e-3, 'disabled', 'heymsfield10_particles', -99.0),
('rwc_q', -99.0, 1, -99.0, -99.0, -99.0, -99.0, -99.0, 3, 50, 'exp', 0.22, 2.2, -99.0, -99.0, 0.00012, 0.006, 'disabled', 'khvorostyanov01_drops', -99.0),
('swc_q', 1.0, -1, -99.0, 0.069, 2.0, 0.3971, 1.88, 3, 50, 'exp', 2.0e06, 0., -99.0, -99.0, 2.0e-04, 0.02, 'disabled', 'heymsfield10_particles', -99.0),
('gwc_q', -99.,-1, -99.0, 169.6, 3.1, -99.0, -99.0, 3, 50, 'exp', -99.0, -99.0, 4.0e6, -99.0, 1.0e-10, 1.0e-2, 'disabled', 'khvorostyanov01_spheres', -99.0)],
dtype=[('hydro_name', 'S15'), ('as_ratio', '<f8'), ('liq_ice', '<i8'), ('rho_ms', '<f8'), ('a_ms', '<f8'), ('b_ms', '<f8'), ('alpha_as', '<f8'), ('beta_as', '<f8'), ('moment_in', '<i8'), ('nbin', '<i8'), ('dist_name', 'S15'), ('p_1', '<f8'), ('p_2', '<f8'), ('p_3', '<f8'), ('p_4', '<f8'), ('d_1', '<f8'), ('d_2', '<f8'), ('scat_name', 'S20'), ('vel_size_mod', 'S30'), ('canting', '<f8')]
)

Hydro_on descriptor file:
descriptorFile = np.array([
#['hydro_name' 'as_ratio' 'liq_ice' 'rho_ms' 'a_ms' 'b_ms' 'alpha_as' 'beta_as' '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', -99.0, 1, -99.0, -99.0, -99.0, -99.0, -99.0, 3, 1, 'mono', -99.0, -99.0, 2.0, 1.0, 2.0e-6, 8.0e-5, 'mie-sphere', 'khvorostyanov01_drops', -99.0),
('iwc_q', 1.0, -1, 700.0, -99.0, -99.0, -99.0, -99.0, 3, 1, 'mono', -99.0, -99.0, 1.564, 0.8547, 1.744e-5, 9.369e-3, 'mie-sphere', 'heymsfield10_particles', -99.0),
('rwc_q', -99.0, 1, -99.0, -99.0, -99.0, -99.0, -99.0, 3, 50, 'exp', 0.22, 2.2, -99.0, -99.0, 0.00012, 0.006, 'mie-sphere', 'khvorostyanov01_drops', -99.0),
('swc_q', 1.0, -1, -99.0, 0.069, 2.0, 0.3971, 1.88, 3, 50, 'exp', 2.0e06, 0., -99.0, -99.0, 2.0e-04, 0.02, 'mie-sphere', 'heymsfield10_particles', -99.0)],
dtype=[('hydro_name', 'S15'), ('as_ratio', '<f8'), ('liq_ice', '<i8'), ('rho_ms', '<f8'), ('a_ms', '<f8'), ('b_ms', '<f8'), ('alpha_as', '<f8'), ('beta_as', '<f8'), ('moment_in', '<i8'), ('nbin', '<i8'), ('dist_name', 'S15'), ('p_1', '<f8'), ('p_2', '<f8'), ('p_3', '<f8'), ('p_4', '<f8'), ('d_1', '<f8'), ('d_2', '<f8'), ('scat_name', 'S20'), ('vel_size_mod', 'S30'), ('canting', '<f8')]
)
I am trying to understand how to set properly the descriptor file to reproduce cloudy/rainy days by simulations.
Thank you in advance for any help in this.
Best regards,
Davide R.

[mariomech]

Hej Davide,

I do simulations based on ERA5. To my knowledge, the microphysics in ERA5 is the same as in the IFS of ECMWF. For that, I use the attached descriptor file. Descriptor files should reflect the microphysics of the model from which the input data for the simulations is taken. Or, if applied to simulations, provide a reasonable description of the knowledge on the hydrometeors, their properties, and size distributions.

There is no dedicated importer for ERA5 included in PAMTRA. I have created one for my use cases that sets up the atmosphere for PAMTRA based on ERA5. However, based on the performance of your clearsky simulations, I do not see an issue with your handling of the ERA5 data. What I'm a bit surprised by is the slightly increased clear

descriptor_file_ecmwf.txt

sky brightness temperatures between 11:00 and 22:00. Or are there any hydrometeors in ERA5 during that time?

Nice to see that you are still using PAMTRA.

All the best

Mario

[Davide88]

Ciao Mario,
thank you for you quick reply!
Yes, I am still using Pamtra and there are still many things I want to explore actually. I think it is a very usefull tool in my field of research, so I am pushing my knowledge of the tool as far as I can.

Thank you for sharing your importer, which I used. I got only minor changes with respect to the previous one, in the order of .1 for some hours.

I checked 11AM .lay generated from the ERA5 file and I noticed there are some hydrometeors in between 755m and 3000m asl. I am using model levels from ERA5, which count approximately 137 vertical levels which I cut beyond a certain altitude and linked to the actual ground level by values extrapolation. I assume that because of the horizontal resolution is relatively low (roughly 28x28km) with respect a upward looking radiometer, there is the possibility that the ERA5 reanalysis for such period includes a cloud fraction, thus possibly hydrometeors, different than 0. I think that 28km is a wide area to cover, thus there is the possibility that the cloud coverage doesn't match very well with the radiometer-detected one, leading to a misaligment of the results when activating hydrometeors calculations. In fact, for other days, I also simulated cases where Pamtra Tb is very lower than the measured one, meaning ERA5 is not including hydrometeors which are actually present. This issue could be solved or at list diminished with higher resolution reanalysis or dedicated high resolution simulations. I tried to explore CERRA, but there is not the possibility to select a limited area, so I should download the whole Europe for each hour, which is not computationally affordable. Moreover, CERRA model levels don't include all the hydrometeors ERA5 model levels do incude. I don't know if you have any suggestion for this based on your experience.
As an example,

ERA5_20250103_1100.txt

I attach the .lay I generated starting from ERA5 for 11AM: you can see how hydrometeors are not 0 between 755 and 3000m asl, as previuously said. I also tryied to force to 0 the values in the whole fifth column, and so I got a good agreement with the measured Tb.
Again thank you for your support.
Have a nice day,
Davide R.