Re: error with rcm2rgrid for WRF interpolation

Erik,

The 1st rule of data processing is to look at your data.

That is why I suggest using printVarSummary(...)

After you used

  rain=dim_avg(rearrange_rain)
  printVarSummary(rain)

You are seeing

Variable: rain

Number of Dimensions: 2
Dimensions and sizes: [record | 29 ] x [west_east | 21 ]

Then you are inputting this to a function that requires
the 2 rightmost dimensions to be lat2d, lon2d

The error message that you are getting tells you that

> fatal:rcm2rgrid: The rightmost dimensions of fi must be nlat2d x
> nlon2d, where nlat2d and nlon2d are the dimensions of the lat2d/lon2d
> arrays

Your "fi" is [record | 29 ] x [west_east | 21 ]

===
I think this 'thread' should taken offline.
I do not want ncl-talk to have multiple messages on this.
There are other issue with your approach.

D

Erik Noble wrote:
> I am still trying to use the rcm2 grid function correctly. I am
> getting this error, which I don't understand.
>
> "fatal:rcm2rgrid: The rightmost dimensions of fi must be nlat2d x
> nlon2d, where nlat2d and nlon2d are the dimensions of the lat2d/lon2d
> arrays
> fatal:Execute: Error occurred at or near line 65 in file tester.ncl"
>
> I have everything in place for the command. I don't see that anyone
> has run into this error in the previous support emails. Could I have
> some advice? both my ncl program and terminal history are below.
> Sincerely,
> Erik
>
> bash-3.1$ ncl tester.ncl
> Copyright (C) 1995-2007 - All Rights Reserved
> University Corporation for Atmospheric Research
> NCAR Command Language Version 5.0.0
> The use of this software is governed by a License Agreement.
> See http://www.ncl.ucar.edu/ for more details.
>
>
> Variable: rainfall
> Type: float
> Total Size: 31668 bytes
> 7917 values
> Number of Dimensions: 3
> Dimensions and sizes: [record | 29] x [south_north | 13] x [west_east | 21]
> Coordinates:
> Number Of Attributes: 6
> FieldType : 104
> MemoryOrder : XY
> description : Total Accumulated Precipitation
> units : mm
> stagger :
> coordinates : XLONG XLAT
> (0)
> (0) Total Accumulated Precipitation: min=0 max=362.638
>
>
> Variable: rainfall
> Type: float
> Total Size: 31668 bytes
> 7917 values
> Number of Dimensions: 3
> Dimensions and sizes: [time | 29] x [lat | 13] x [lon | 21]
> Coordinates:
> Number Of Attributes: 6
> FieldType : 104
> MemoryOrder : XY
> description : Total Accumulated Precipitation
> units : mm
> stagger :
> coordinates : XLONG XLAT
>
>
> Variable: rain
> Type: float
> Total Size: 2436 bytes
> 609 values
> Number of Dimensions: 2
> Dimensions and sizes: [time | 29] x [lon | 21]
> Coordinates:
> Number Of Attributes: 6
> coordinates : XLONG XLAT
> stagger :
> units : mm
> description : Total Accumulated Precipitation
> MemoryOrder : XY
> FieldType : 104
> fatal:syntax error: procedure printMinMax expects 2 arguments, got 1
> fatal:error at line 34 in file tester.ncl
>
>
>
> Variable: lat2d
> Type: float
> Total Size: 1092 bytes
> 273 values
> Number of Dimensions: 2
> Dimensions and sizes: [south_north | 13] x [west_east | 21]
> Coordinates:
> Number Of Attributes: 5
> FieldType : 104
> MemoryOrder : XY
> description : LATITUDE, SOUTH IS NEGATIVE
> units : degrees_north
> stagger :
>
>
> Variable: lon2d
> Type: float
> Total Size: 1092 bytes
> 273 values
> Number of Dimensions: 2
> Dimensions and sizes: [south_north | 13] x [west_east | 21]
> Coordinates:
> Number Of Attributes: 5
> FieldType : 104
> MemoryOrder : XY
> description : LONGITUDE, WEST IS NEGATIVE
> units : degrees_east
> stagger :
>
>
> Variable: mlon
> Type: integer
> Total Size: 4 bytes
> 1 values
> Number of Dimensions: 1
> Dimensions and sizes: [1]
> Coordinates:
> (0) 22
>
>
> Variable: nlat
> Type: integer
> Total Size: 4 bytes
> 1 values
> Number of Dimensions: 1
> Dimensions and sizes: [1]
> Coordinates:
> (0) 14
> fatal:rcm2rgrid: The rightmost dimensions of fi must be nlat2d x
> nlon2d, where nlat2d and nlon2d are the dimensions of the lat2d/lon2d
> arrays
> fatal:Execute: Error occurred at or near line 65 in file tester.ncl
>
> fatal:Variable (raingrd) is undefined
> fatal:Execute: Error occurred at or near line 66 in file tester.ncl
>
> bash-3.1$ cat tester.ncl
> ;rrPlot Hovmoller of WRF and TRMM Accumulated Daily Rain
> ;Create the WRF data from WRF output as .nc file
> ;Create the TRMM as an hovmoller ascii file from
> http://lake.nascom.nasa.gov/Giovanni/tovas/
> ; Remove headers from ascii file; also figure out number of rows
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl"
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_csm.ncl"
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/contributed.ncl"
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/shea_util.ncl"
> load "$NCARG_ROOT/lib/ncarg/nclscripts/wrf/WRFUserARW.ncl"
>
> ; Global
> dir = "./"
> wrfoutfile="Region_1_daily_24hr_wrf_West_Africa_2006-09.nc"
> obsfile = "TRMMR1_9-2_10-1.ascii"
> region = "Region_1"
> regiontitle = "Atlantic ITCZ"
> elaptime = 29 ; normally 30, but missing a day
> ; Read in 1st file
> a = addfile (wrfoutfile,"r")
> Rain = a->RAINC
> rainfall = a->RAINC
> printVarSummary(rainfall) ; look at the variable
> printMinMax(rainfall, True ) ; contributed.ncl
> rainfall!0 = "time"
> rainfall!1 = "lat"
> rainfall!2 = "lon"
> printVarSummary(rainfall)
>
> rearrange_rain = rainfall(time|:,lon|:,lat|:) ; reorder coordinates
> for next step
> ; Average all lats as a function of lon for Time vs. Longitude plot
> (Hovmueller diagram)
> rain=dim_avg(rearrange_rain) ; avg rightmost dimension with center dimension
> copy_VarMeta(rearrange_rain, rain)
> printVarSummary(rain)
> printMinMax(rain)
>
> ;WRF XLAT/XLONG are (Time,south_north,west_east)
> ;Generally, they do not change with time.
> ;Hence, read then as 2D arrays
>
> ; These are [*][*] (two dimensional)
> lat2d = a->XLAT(0,:,:)
> lon2d = a->XLONG(0,:,:)
> lat2d_at_units = "degrees_north" ; not required for interpolation
> lon2d_at_units = "degrees_east"
> printVarSummary(lat2d)
> printVarSummary(lon2d)
> ;Your 'target' (rectilinear) grid is lat[*],lon[*].
> ;This could be from some other model
> ;or you could manually create via
> latS = 5 ; regular grid output
> latN = 11
> lonL = -30
> lonR = -20
>
> mlon = 22 ; however many points you want, here 22
> because 0.5 degree
> lon = fspan (lonL, lonR, mlon)
> lon!0= "lon"
> lon_at_units = "degrees_east"
> print(mlon)
> nlat = 14 ; however many points you want
> lat = fspan (latS,latN, nlat)
> lat!0= "lat"
> lat_at_units = "degrees_north"
> print(nlat)
> raingrd = rcm2rgrid(lat2d,lon2d,rain,lat,lon,0)
> printVarSummary(raingrd)
> ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
>
>
>
> On Sun, Jan 18, 2009 at 5:16 AM, Erik Noble <enoble_at_giss.nasa.gov> wrote:
>> Dear Dennis. Thank you. Your help was more than expected help. Yes
>> .....I got the latitude/longitude labels backwards...:-| Thank you.
>>
>> On Sun, Jan 18, 2009 at 3:54 AM, Dennis Shea <shea_at_ucar.edu> wrote:
>>> a = addfile("WRF.nc","r")
>>> fi = a->T ; (Time,bottom_top,south_north,west_east)
>>> printVarSummary(fi) ; look at the variable
>>> printMinMax(fi, True ) ; contributed.ncl
>>>
>>> WRF XLAT/XLONG are (Time,south_north,west_east)
>>> Generally, they do not change with time.
>>> Hence, read then as 2D arrays
>>>
>>> ; These are [*][*] (two dimensional)
>>> lat2d = a->XLAT(0,:,:)
>>> lon2d = a->XLONG(0,:,:)
>>> lat2d_at_units = "degrees_north" ; not required for interpolation
>>> lon2d_at_units = "degrees_east"
>>>
>>> Your 'target' (rectilinear) grid is lat[*],lon[*].
>>> This could be from some other model
>>> or you could manually create via
>>>
>>> latS =-20 ; regular grid output
>>> latN = 35
>>> lonL =-35
>>> lonR = 35
>>>
>>> mlon = 50 ; however many points you want
>>> lon = fspan (lonL, lonR, mlon)
>>> lon!0= "lon"
>>> lon_at_units = "degrees_east"
>>>
>>> nlat = 40 ; however many points you want
>>> lat = fspan (latS,latN, nlat)
>>> lat!0= "lat"
>>> lat_at_units = "degrees_north"
>>>
>>> -----
>>> v5.1.0 has a faster code but the interface is exactly the same
>>>
>>> Erik Noble wrote:
>>>> Thank you Dennis, for pointing me to the rcm2rgrid instructions.
>>>> Although there are examples, I'm not clear from the examples on what
>>>> to put for the "lat" and "lon" arguments. It says that they are 1D
>>>> arrays that specify the lat and lon coordinates of the regular grid
>>>> and that they should be monotonically increasing. My model output is
>>>> regional ( latitudes 35W to 35E, longitudes 20S to 35N).
>>> ?latitudes with W and E ... no
>>> ?longitudes with N and S ... no
>>>
>>> This is confusing
>>>
>>> Would I put
>>>> these values in for the "lat" and "lon" arguments respectively(
>>>> {-35:35}",{-20:35})?
>>>>
>>>> function rcm2rgrid (
>>>> lat2d [*][*] : numeric,
>>>> lon2d [*][*] : numeric,
>>>> fi : numeric,
>>>> lat [*] : numeric,
>>>> lon [*] : numeric,
>>>> Option : numeric
>>>> )
>>>>
>>>> return_val : numeric
>>>> Thank you for your help.
>>>> -Erik
>>>> On Sat, Jan 17, 2009 at 1:13 PM, Dennis Shea <shea_at_ucar.edu> wrote:
>>>>> http://www.ncl.ucar.edu/Document/Functions/Built-in/rcm2rgrid.shtml
>>>>>
>>>>> WRF data have multidimensional lat/lon arrays:
>>>>> XLAT[*][*], XLONG[*][*]
>>>>>
>>>>> linint2 had coordinates prototypes as one dimensional:
>>>>> lat[*], lon[*]
>>>>>
>>>>> The rcm2rgrid in version 5.1.0 is more efficient.
>>>>>
>>>>>
>>>>>
>>>>> Erik Noble wrote:
>>>>>> Hi. I am trying to learn how to use NCL to interpolate WRF rain data
>>>>>> to observed precipitation from Satellite (TRMM).
>>>>>> Some say use the linint2 function to bilinearly interpolate.
>>>>>> Some don't:
>>>>>> http://www.ncl.ucar.edu/Support/talk_archives/2008/0484.html
>>>>>>
>>>>>> Do we have any examples of using linint2 with wrf data?
>>>>>> -Erik
>>>>>> _______________________________________________
>>>>>> ncl-talk mailing list
>>>>>> List instructions, subscriber options, unsubscribe:
>>>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>>>
>>>>>
>>>
>>>
>>>
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Sun Jan 18 2009 - 08:40:04 MST