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

geogrid bad interpolation along coastline

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.

davidwrf

Member
Hi,

I have downloaded and compiled the most recent WRF and WPS versions.
I'm running geogrid.exe with the default albedo options:
*****************************************
name=ALBEDO12M
priority=1
dest_type=continuous
z_dim_name=month
masked = water
fill_missing = 8.
interp_option=default:four_pt+average_4pt+average_16pt+search
rel_path=default:albedo_modis/
**********************************************

The most fine nest is of 0.5km resolution.
The problem is that over land, along the coastline, the albedo map looks very coarse. Further inland, it looks much finer.
I assume that the reason is that when interpolating for land grid cells along the coastline, the few points over water (with value 8) contribute substantially to the outcome of the albedo value. Is that correct?

If so, is there way to tell NOT to use the water grid cells when evaluating the albedo over land?

David
 
That the ALBEDO12M field looks coarser near land-sea boundaries sounds more like evidence that the geogrid program is not using water values. The 'index' file for the 'albedo_modis' dataset contains the following:
Code:
type=continuous
projection=regular_ll
dx=0.05
dy=-0.05
known_x=1.0
known_y=1.0
known_lat=89.975
known_lon=-179.975
wordsize=2
tile_x=1200
tile_y=1200
tile_z_start=1
tile_z_end=12
tile_bdr=3
scale_factor=0.01
signed = yes
missing_value=-999
units="percent"
description="Monthly MODIS surface albedo"
In particular, there is a missing value (-999) specified. I haven't actually checked to verify that missing values are indeed present in the dataset, but if they are (and, presumably, they are used for water points), then the geogrid program should be avoiding the use of any water values in its interpolation. Note that the entries
Code:
        masked = water
        fill_missing = 8.
in the ALBEDO12M section of the GEOGRID.TBL file instruct the geogrid program to not interpolate to grid cells that are defined as water in the high-resolution WRF landmask field, and to instead set the ALBEDO12M field for those grid cells to a value of 8; these entries do not influence how the geogrid program interprets the source dataset (albedo_modis, in this case).

If you could post your namelist.wps file, that would allow us to try reproducing what you're seeing and to hopefully provide some more specific recommendations.
 
Hi,
Thanks for the quick reply.
Below is the namelist.wps used (with the original GEOGRID.TBL file).
You'll be able to see the coarser ALBEDO field along the coastline in the most finer domain.

**********************************************
&share
wrf_core = 'ARW',
max_dom = 4,
start_date = '2013-07-12_18:00:00','2013-07-12_18:00:00','2013-07-12_18:00:00','2013-07-12_18:00:00',
end_date = '2013-07-14_12:00:00','2013-07-14_12:00:00','2013-07-14_12:00:00','2013-07-14_12:00:00',
interval_seconds = 21600
io_form_geogrid = 2,
/

&geogrid
parent_id = 1, 1, 2, 3,
parent_grid_ratio = 1, 3, 3, 3,
i_parent_start = 1, 63, 60, 29,
j_parent_start = 1, 62, 60, 50,
e_we = 160, 160, 130, 130,
e_sn = 172, 193, 127, 133,
geog_data_res = '5m_mds_15s', '2m_mds_15s', '30s_mds_15s', '15s_mds_15s',
dx = 13500,
dy = 13500,
map_proj = 'lambert',
ref_lat = 31.70,
ref_lon = 33.70,
truelat1 = 30.0,
truelat2 = 60.0,
stand_lon = 35.0,
geog_data_path = '/ssdhome/dudi/Build_WRF/WPS_GEOG'
/

&ungrib
out_format = 'WPS',
prefix = 'FILE',
/

&metgrid
fg_name = 'FILE',
io_form_metgrid = 2,
/
**********************************************

David
 
Top