Noah-MP Snow Initialization

[politeness]

Hi, I am interested in modifying the snow water equivalent and snow depth fields in my WRF restart file (i.e., adding snow to a snow-free grid). I am wondering what other fields I will need to modify for physical consistency. For example, upon initialization will Noah-MP determine the appropriate number and thickness of snow layers and set snow cover=1?

Thank you in advance for any insight you can provide.

 

[Ming Chen]

Hi,
If you look at wrfrst file, you will find the following data related to snow:

       float ACSNOW(Time, south_north, west_east) ;
                ACSNOW:FieldType = 104 ;
                ACSNOW:MemoryOrder = "XY " ;
                ACSNOW:description = "ACCUMULATED SNOW" ;
                ACSNOW:units = "kg m-2" ;
                ACSNOW:stagger = "" ;
                ACSNOW:coordinates = "XLONG XLAT XTIME" ;
                ACSNOM:description = "ACCUMULATED MELTED SNOW" ;
        float SNOW(Time, south_north, west_east) ;
                SNOW:FieldType = 104 ;
                SNOW:MemoryOrder = "XY " ;
                SNOW:description = "SNOW WATER EQUIVALENT" ;
                SNOW:units = "kg m-2" ;
                SNOW:stagger = "" ;
                SNOW:coordinates = "XLONG XLAT XTIME" ;
        float SNOWH(Time, south_north, west_east) ;
                SNOWH:FieldType = 104 ;
                SNOWH:MemoryOrder = "XY " ;
                SNOWH:description = "PHYSICAL SNOW DEPTH" ;
                SNOWH:units = "m" ;
                SNOWH:stagger = "" ;
                SNOWH:coordinates = "XLONG XLAT XTIME" ;
        float SNOWNC(Time, south_north, west_east) ;
                SNOWNC:FieldType = 104 ;
                SNOWNC:MemoryOrder = "XY " ;
                SNOWNC:description = "ACCUMULATED TOTAL GRID SCALE SNOW AND ICE" ;
                SNOWNC:units = "mm" ;
                SNOWNC:stagger = "" ;
                SNOWNC:coordinates = "XLONG XLAT XTIME" ;
        float SNOWNCV(Time, south_north, west_east) ;
                SNOWNCV:FieldType = 104 ;
                SNOWNCV:MemoryOrder = "XY " ;
                SNOWNCV:description = "TIME-STEP NONCONVECTIVE SNOW AND ICE" ;
                SNOWNCV:units = "mm" ;
                SNOWNCV:stagger = "" ;
                SNOWNCV:coordinates = "XLONG XLAT XTIME" ;
                TSNAV:description = "AVERAGE SNOW TEMPERATURE" ;
        float SNOWC(Time, south_north, west_east) ;
                SNOWC:FieldType = 104 ;
                SNOWC:MemoryOrder = "XY " ;
                SNOWC:description = "FLAG INDICATING SNOW COVERAGE (1 FOR SNOW COVER)" ;
                SNOWC:units = "" ;
                SNOWC:stagger = "" ;
                SNOWC:coordinates = "XLONG XLAT XTIME" ;
       float TSNAV(Time, south_north, west_east) ;
                TSNAV:FieldType = 104 ;
                TSNAV:MemoryOrder = "XY " ;
                TSNAV:description = "AVERAGE SNOW TEMPERATURE" ;
                TSNAV:units = "C" ;
                TSNAV:stagger = "" ;
                TSNAV:coordinates = "XLONG XLAT XTIME" ;
        float SNOALB(Time, south_north, west_east) ;
                SNOALB:FieldType = 104 ;
                SNOALB:MemoryOrder = "XY " ;
                SNOALB:description = "ANNUAL MAX SNOW ALBEDO IN FRACTION" ;
                SNOALB:units = "" ;
                SNOALB:stagger = "" ;
                SNOALB:coordinates = "XLONG XLAT XTIME" ;
       float ACSNOM(Time, south_north, west_east) ;
                ACSNOM:FieldType = 104 ;
                ACSNOM:MemoryOrder = "XY " ;
                ACSNOM:description = "ACCUMULATED MELTED SNOW" ;
                ACSNOM:units = "kg m-2" ;
                ACSNOM:stagger = "" ;
                ACSNOM:coordinates = "XLONG XLAT XTIME" ;

These variables should be changed at the grids where you will make changes.

 

[politeness]

Ming,
Thank you for your response. Relatedly, I wonder if you can explain cases that I am seeing in the WRF restart file where ISNOW=0, but 0<SNOWH<0.15 AND SNOWC==1. How does it happen that a snow-covered grid with a depth of snow does not have a layer of snow characterized?

Thanks,
Craig

 

[Ming Chen]

Craig,
My understanding is that, NoahMP always specifies snow layer when it starts from initial run or from restart file.If you look at the code " module_sf_noahmplsm.F", you will find the following piece of codes:

! snow/soil layer thickness (m)

     DO IZ = ISNOW+1, NSOIL 
         IF(IZ == ISNOW+1) THEN
           DZSNSO(IZ) = - ZSNSO(IZ)
         ELSE  
           DZSNSO(IZ) = ZSNSO(IZ-1) - ZSNSO(IZ)
         END IF
     END DO

Even if isnow=0 in wrfrst files, snow layers are specified.

 

[politeness]

Hi Ming,
Sorry, but the code snippet doesn't help with the issue I am seeing. I think the explanation likely lies in the output/update sequence of isnow vs. SWE and SNOWH. For example, in my restart file, there is a grid with the following attributes:

snowc=1.0
isnow=0
tsno=[0,0,0]
zsnso=[0,0,0,-0.1, -0.4, -1.0, -2.0]
snliq=[0,0,0]
snice=[0,0,0]
snowh=0.0506
snow=10.117

It seems to me that either snowh and snow are from a previous or later timestep than the column variables (i.e., zsnso, tsno, snliq, snice, isnow).

Thanks,

 

[Ming Chen]

Hi, Craig,
ISNOW is a variable specifically used in SSIB. It has nothing to do with snow fields in NoahMP.
snowh and snow in NoahMP are from a previous timestep.
Does this answer your question?