W2W vs WRF/WPS interpolation of LCZ values

[jesusff]

From WRFcoordination created by jesusff: FPS-URB-RCC/WRFcoordination#1

https://github.com/FPS-URB-RCC/WRFcoordination/blob/c7f3201f515edac779f5c68c052c6bd18ea94d4b/URBPARM_LCZ.TBL#L598-L603

Lluís: Seems to me it should be 6?
Oscar: It should be 2. We have put 5 a while ago with Alberto because BEP only accepted levels of 5 m in the past.

[jesusff]

Comment by Matthias Demuzere

I can confirm.

I m not sure what the original question is, but if you use W2W then these values are set by default.

In addition: the next release of WRF/WPS will allow you to use LCZs without having to use W2W. All of the required pre-processing will be done within WRF/WPS. It is for that reason I created the hybrid Copernicus-Modis-LCZ land use map.

Best,
Matthias

[jesusff]

Thank you for jumping in, @matthiasdemuzere

if you use W2W then these values are set by default.

Do you mean that W2W has hard-coded values for e.g. building heights, regardless of the setting in URBPARM_LCZ.TBL? We will likely keep using your tool in the FPS

the next release of WRF/WPS will allow you to use LCZs without having to use W2W

Do you mean 4.5 (released a few days ago)? or is it intended for 4.5.1? 4.6? There seems to be already some capability to process your new data in v4.5

[jesusff]

comment by Andrea Zonato

Hi all,

In version 4.5 you can find the LCZ map at 100m resolution. However, regarding the interpolation method to create point-by-point values has not been released yet, I'm working on it!

[jesusff]

@andreazonato, is this w2w interpolation method planned for the upcoming WPS v4.6?

[LluisFB]

Regarding to that, be aware.
in the geo_em.d[nn].nc when using LCZ, a variable called URB_PARAM is created with 131 values per grid point. This variable stores all the parameters describing the characteristics of the urban grid points (see their meaning from this NUDAPT document). The values found during running time in this variable, are the ones being used by the WRF model.
Because W2W fills URB_PARAM with values at certain positions, be aware that any modification in the URBPARAM_LCZ.TBL will not be taken by WRF, if URB_PARAM has values at the given parameter. This is the case at least for: distribution of heights of the buildings, ...

[jesusff]

Thank you, Lluís.

@AndresSimon was also looking into this lately. This is related to #7, as the file developed there might not be used if we keep using w2w

[AndresSimon]

I realized that even though the variables inside URB_PARAM reach the wrfinput_d02 file as BUILD_AREA_FRACTION, BUILD_HEIGHT and BUILD_SURF_RATIO, they are not considered by BEP, which ends up using the values in URBPARM_LCZ.TBL.
@andreazonato, any clue about this? Is there any code modification needed?
Thanks!!

[andreazonato]

Hi Andres.

That's pretty strange since it shall read them inside BEP-BEM and overwrite them once inside BEP-BEM.

Can you put a print inside to check wether the values are true?

[AndresSimon]

Hi Andrea,

Where in BEP-BEM are read?
I am printing lp_urb2d and hgt_urb2d, right before the BEP1D call and they are zero.
However, FRC_URB2D takes the value provided by w2w.
On the other hand ws and bs take the same value as in URBPARM_LCZ.TBL

[andreazonato]

That means that they are 0. And that is strange. And you said they are correctly defined in the wrfinput right?

[AndresSimon]

Yes, the values in wrfinput make sense.
Where are those parameters read in BEP-BEM? Because I don't find it...

[andreazonato]

call icBEPHI_XY(iurb,hb_u,hi_urb1D,ss_urb,pb_urb, &
nz_urb(iurb),z_u)

Here! If values are present, LCZ values are replaced!

[LluisFB]

Sorry, I am not sure if this is the case.
We performed (my PhD student) sensitivity simulations using BEP+BEM where we modified the values in the URBPARAM_LCZ.TBL. When we checked the results, we did not obtain any sensitivity. This make us realize that modifications where not been taken into consideration. When we zeroed the values from the geo_em file, then we obtain the sensitivity