Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

There is a Huge Difference between Amount of Rain Output and Rain Input

This post was from a previous version of the WRF&MPAS-A Support Forum. New replies have been disabled and if you have follow up questions related to this post, then please start a new thread from the forum home page.


New member

I've been working on rain forecasting. However, there is huge difference between output and input data which is GFS 0.25 degree resolution forecast data.

The amount of rain output is higher than GFS data generally. For example, when rain variable of input data is 0.1 mm/hr, WRF gives 30 mm/hr. There is no high amount of rain such 30 mm/hr in same location in the reality by the way. I did this comparison for ERA5 reanalysis data too, and there was similar result.

I know that "RAINNC" variable is accumulated data. I made calculations considering this when comparing with other data. I know that is not the problem but, I can't find the solution why this is happening either. I am giving another information about my model design. I would be very happy if you could enlighten me on this matter. :idea: :ugeek:

Model resolutions:
Spatial resolution = 27 km, 9 km and 3 km (nest)
Time resolution = 75 hr (48 hr is spin-up time)
Vertical level = 50

Physics Options:
mp_physics = 6, 6, 6,
ra_lw_physics = 4, 4, 4,
ra_sw_physics = 4, 4, 4,
radt = 10, 10, 10,
sf_sfclay_physics = 1, 1, 1,
sf_surface_physics = 2, 2, 2,
bl_pbl_physics = 1, 1, 1,
bldt = 0, 0, 0,
cu_physics = 0, 0, 0,
cudt = 5, 5, 5,
icloud = 1,
surface_input_source = 1,
num_land_cat = 21,
num_soil_layers = 4,
sf_urban_physics = 0, 0, 0,
hailcast_opt = 1, 1, 1,
I can confirm that it works, I stumbled upon this issue earlier on and managed to solve it by myself. I'd like to know exactly why/how this happens as not to get similar issues, but at least, it works.
For 27km and 9km resolutions, we need to turn on cumulus scheme to describe convective processes. Otherwise the physics will be incomplete and lead to wrong results. This is why the model crashed.
I have a similar problem.
I have two domains with 18 and 6 km resolution.
After 10 hours of simulation precipitation field in the nested domain looks unrealistic.
In particular, there are two unrealistic local maxima >120 mm over the sea.

Precipitation field in the parent domain looks normal.

I tried different mp_physics and cu_physics options but it didn't help.
What causes such intense precipitation maxima in the nested domain?

I use GFS as initial and boundary conditions. Cumulus parametrization is turned on in both domains.
Here is my namelist.input.
time_step = 72,
max_dom = 2,
dx = 18000, 6000,
dy = 18000, 6000,
e_vert = 38, 38,
mp_physics = 16, 16,
ra_lw_physics = 4, 4,
ra_sw_physics = 4, 4,
radt = 18, 6,
sf_surface_physics = 2, 2,
sf_sfclay_physics = 1, 1,
bl_pbl_physics = 1, 1,
bldt = 0, 0,
cu_physics = 1, 1,
cudt = 5, 5,
icloud = 1,
surface_input_source = 1,
num_soil_layers = 4,
sf_urban_physics = 0, 0,
num_land_cat = 21,
6 km resolution is located within the typical 'grey zone', i.e., for model resolutions between 4-10km, it is hard to describe convective precipitation. This is because almost all the cumulus schemes are designed for coarse resolution cases (grid interval > 10km), while the cloud-resolving resolution must be be 3-4km or higher. This is why 6km resolution cases often yield unreasonable results.
I would suggest that you turn off cumulus scheme for the 6km domain, but keep it turned on for the 18km domain. Then rerun the case and hope the results could be better.
Thank you for the reply.
I turned off cumulus parametrization in the nested domain but it didn't help.
It seems, that in my case the model can't reproduce precipitation field with high resolution.
I am going to use the precipitation field from 18 km domain.