Project

General

Profile

Wiki » Snowpack model CROCUS » Development in progress »

Explicit representation of Light-Absorbing Particles

In order to use the model configuration including Light-Absorbing Particles (LAP) in snow presented in Tuzet et al. (2017) you have to install the branch cen_dev from the SURFEX_Git2 repository.

To activate the explicit representation of LAP as well as TARTES radiative transfer scheme, you only have to make some changes in the default namelist.
These changes are detailed hereafter for each block of the namelist. A default namelist activating the explicit representation of LAP is also attached to this wiki page.
It is noteworthy that the s2m command to launch is the same as a classic Crocus run, only the namelist (and eventually the atmospheric forcing file) must be modified.

Necessary changes in namelist

&NAM_PREP_ISBA_SNOW

NIMPUR=0/1/2              ; default=0

NIMPUR is the number of LAP types you want to use in your simulation.
For now you can set NIMPUR=1 if you want to run simulations with BC only and NIMPUR=2 if you want to run, simulations with BC and Dust.
If you want to run a simulation with dust only you can set NIMPUR=2 and prescribe no BC deposition.

&NAM_IO_OFFLINE

LSPECSNOW = .TRUE.        ; default=.FALSE. ,   #Enable spectral computation inside SURFEX/Crocus, necessary to run TARTES radiative transfer scheme    
NIMPUROF=0/1/2            ; default=0       #Initialize the number of LAP types in the OFFLINE module, you have to set NIMPUROF to the same value as NIMPUR.

&NAM_DIAG_ISBAn

LPROBANDS= .TRUE.         ; default=.FALSE.  #Enable the spectral resolution of Crocus diagnostics, necessary if you want to get spectral albedo and spectral direct/diffuse ratio diagnostics

&NAM_ISBA_SNOWn

CSNOWRAD='T17'            ; default=B92   #Set the radiative transfer scheme to TARTES+LAP as in Tuzet et al. (2017)     
#The following option is currently unavailable, setting it to .TRUE. will cause some errors 
#LATMORAD=.FALSE.          ; default=.FALSE. # This option is not stable yet, but it is supposed to compute the direct/diffuse ratio directly from atmospheric informations (AOD, Ozone column, Water column..)

&NAM_WRITE_DIAG_SURFn

CSELECT: You have to add 'SNOWIMP1','SNOWIMP2','SPEC_ALB', 'DIFF_RATIO', SPEC_TOT in this field.

CSELECT makes a selection of the diagnostics you want to compute in your output NetCDF file (PRO file).
SNOWIMP1 is the BC concentration in each snow layer (g/g_of_snow)
SNOWIMP2 is the Dust concentration in each snow layer (g/gg_of_snow)
SPEC_ALB is the spectral albedo for the 186 Crocus spectral bands (from 300 to 4000nm included by step of 20nm; i.e [300,320...,3980,4000])
DIFF_RATIO is the spectral direct to diffuse ratio for the 186 Crocus spectral bands (from 300 to 4000nm included by step of 20nm; i.e [300,320...,3980,4000])
SPEC_TOT is the total incoming irradiance after spectral repartition i.e the spectral irradiance used by TARTES for radiative transfer computations

Optional changes (to activate LAP deposition)

The following changes set the way you want to prescribe your deposition fluxes in input of SURFEX/Crocus.
You can either feed the model with prescribed and constant deposition fluxes or introduce a wet and dry deposition field directly in the forcing file.

  • If you want to prescribe constant deposition fluxes over time:

&NAM_SURF_SNOW_CSTS

XIMPUR_WET(1)=5.e-11     ; default=0.
XIMPUR_WET(2)=5.e-9      ; default=0.
XIMPUR_DRY(1)=1.e-11     ; default=0.
XIMPUR_DRY(2)=1.e-9     ; default=0.

BE CAREFULL THERE IS A BUG IN SOME VERSION OF FORTRAN WHEN SETTING A NAMELIST VARIABLE TO 1.E-10.
These variable set the different depositions fluxes and are expressed in g m^(-2) s^(-1).

XIMPUR_WET corresponds to the initial amount of LAP present in the falling snow (wet deposition) for each impurity type, activated in case of rain or snow.
XIMPUR_DRY corresponds to the dry deposition coefficient always activated.
You can precise the type of LAP into parenthesis as done above. e.g XIMPUR_WET(1)=1.e-9 set the wet depostion coefficient of BC to 1.e-9 g m^(-2) s^(-1)

  • If you want to prescribe directly the deposition fluxes at each model time step from a forcing file:

&NAM_IO_OFFLINE

LFORCIMP = .TRUE.         ; default=.FALSE.
#The following option is currently unavailable, setting it to .TRUE. will cause some errors
#LFORCATMOTARTES =.FALSE.  ; default=.FALSE. # Activate the prescription of aerosol optical depth and ozone column from forcing file if LATMORAD=.TRUE.

When you activate LFORCIMP by setting it to .TRUE., you have to add new variables in your forcing file.
You have two variables to add for each type of LAP: IMPWET (wet deposition coefficient) and IMPDRY (dry deposition coefficient). Both these deposition fluxes have to be in g m^(-2) s^(-1).
If you have one type of LAP you will need IMPWET1 and IMPDRY1.
If you have two type of LAP you will need IMPWET1 and IMPDRY1 (BC) and IMPWET2 and IMPDRY2 (Dust). etc...
These new variables must be defined with the same dimension as the snowfall rate for exemple.
If those variables are not defined in your forcing files you will get an error when running the simulation.
Note that if you activate LFORCIMP the deposition values contained in NAM_SURF_SNOW_CSTS are ignored.

##Nominal behaviour of LFORCATMOTARTES when it will be repared##
When you activate LFORCATMOTARTES by setting it to .TRUE., you have to add new variables in your forcing file : AODTOT (total aerosol optical depth) and OZONE (total ozone column).
These new variables must be defined with the same dimension as the snowfall rate for exemple. If those variables are not defined in your forcing files you will get an error when running the simulation. ###

P.S: In the attached namelist the other parameters have default values, you might need to tune this parameters for your study or to use the multiphysics ensemble of Crocus, ESCROC described in Lafaysse et al. (2017).

If you need to add another type of LAP in the model please open a new ticket on the support page and someone will get back to you.

Initialization of an user prescribed snowpack with LAP

In the case you want to initialize an user-prescribed snowpack by defining manually it's layers, it's possible to initialize the concentration of LAPs in each layer by adding the values in NAM_PREP_ISBA_SNOW by adapting the following exemple:

&NAM_PREP_ISBA_SNOW

 XIMPURSNOW(:,1)= 0,0,5e-9,0,0  #BC concentration (g/g_of_snow) in each layer (5 layers here)
 XIMPURSNOW(:,2)= 0,0,5000e-9,0,0 #Dust concentration (g/g_of_snow) in each layer(5 layers here)

And so on if you have more impurity types. By default the concentration in all layers will be 0.

Scavenging of impurities through the snowpack with melt water

As fully explained in the original article (Tuzet et al. 2017), it is possible to activate the scavenging of impurities through the snowpack with mel water.
The representation of water percolation in Crocus are, for now, quite simple and might be affected by large uncertainties so the scavenging of impurities is disabled by default.
If you want to run a simulation including the scavenging of impurities, you have to adapt the following example:

&NAM_PREP_ISBA_SNOW

 XSCAVEN_COEF(1)=0.2 #20% of BC scavenged with percolating water
 XSCAVEN_COEF(2)=0.05 #5% of dust scavenged with percolating water

Note : A scavenging coefficient <1 means that the percolating water is less concentrated in impurities than the layer from which the water is leaving.
This leads to an enrichement of impurities as water is percolating through the snowpack.

Information on default values used in TARTES radiative transfer scheme

The refractive of ice is taken from Warren and Brandt. 2008.

The values of the shape parameters B and g used in Kokhanovsky and Zege(2004) theory are set to 1.6 and 0.845 respectively (according to Libois et al. 2013 and Dumont et al. 2017)

The Mass Absoprtion Efficiency (MAE) of black carbon is based on refractive index advised by Bond and Bergstrom 2006, with a value of 11.25 kg m-1 at 550nm as in Tuzet et al. 2019.
The Mass Absoprtion Efficiency (MAE) of mineral dust is based on refractive index found in Caponi et al. 2017 for Lybian dust of PM 2.5

Important information about new diagnostics

By activating these new developments, you will have new diagnostics in your PRO file (if you correctly modified LSELCT in &NAM_WRITE_DIAG_SURFn).
The LAP concentration diagnostics (SNOWIMP) have classic dimensions (time,number of point,number of snow layers).
The spectral diagnostics (SPEC_ALB,DIFF_TOT and SPEC_TOT) have special dimensions (time,number of spectral bands,number of snow layers).
For now the spectral bands run from 300 to 4000 (included) by step of 20, meaning 186 spectral bands. (300,320,340,......,3980,4000)

Important notes about new developpements since Tuzet et al. (2017)

At the end of the snow season, when scavenging is disabled, the LAP present in melting layers accumulate in the uppermost layers. This is the enrichement of the surface in LAP content, described for instance in Sterle et al. (2013).
In Crocus, this uppermost layer have its own dynamics, and its thickness can strongly vary from one timestep to another. When the mass of LAP in the uppermost layer is important (>1 microgram of black carbon for instance), the thickness of the uppermost layer strongly influences the radiative transfer. The spectral albedo then shows a strong dependence to the thickness of the top layer, and hence to the timestep while the mass of impurity is the same.
This behavior is just a numerical artifact that is not desired in Crocus simulations. To minimize this effect, the mass of LAP in the uppermost layer is equally reparted in the 10 first millimeters of SWE at each timestep.
This way, the impact of the surface mass of LAP on snow radiative transfer does not depend on the numerical thickness of the uppermost layer.

It is noteworthy that the default values of LAP mass absorption efficiency have been modified since the original article.

Reference

Bond, T. C. and Bergstrom, R. W.: Light absorption by carbonaceous particles: An investigative review, Aerosol science and technology, 40,
27–67, 2006

Caponi, L., Formenti, P., Massabo, D., Biagio, C. D., Cazaunau, M., Pangui, E., Chevaillier, S., Landrot, G., Andreae, M. O., Kandler, K.,
et al.: Spectral-and size-resolved mass absorption efficiency of mineral dust aerosols in the shortwave spectrum: a simulation chamber
study, Atmospheric Chemistry and Physics, 17, 7175–7191, 2017.

Dumont, M., Arnaud, L., Picard, G., Libois, Q., Lejeune, Y., Nabat, P., Voisin, D., and Morin, S.: In situ continuous visible and near-infrared
spectroscopy of an alpine snowpack, The Cryosphere, 11, 1091–1110, https://doi.org/10.5194/tc-11-1091-2017, http://www.the-cryosphere.
net/11/1091/2017/, 2017.

Kokhanovsky, A. and Zege, E.: Scattering optics of snow, Applied Optics, 43(7), 1589–1602, https://doi.org/doi:10.1364/AO.43.0001589,
2004.

Sterle, K.M., McConnell, J.R., Dozier, J., Edwards, R. and Flanner, M.G., 2013. Retention and radiative forcing of black carbon in eastern Sierra Nevada snow. The Cryosphere, 7(1), pp.365-374.

Libois, Q., Picard, G., France, J. L., Arnaud, L., Dumont, D., Carmagnola, C. M., and King, M. D.: Influence of grain shape on light penetration
in snow, The Cryosphere, 7, 1803–1818, https://doi.org/10.5194/tc-7-1803-2013, 2013.

Tuzet, F., Dumont, M., Lafaysse, M., Picard, G., Arnaud, L., Voisin, D., Lejeune, Y., Charrois, L., Nabat, P., and Morin, S.: A multilayer
physically based snowpack model simulating direct and indirect radiative impacts of light-absorbing impurities in snow, The Cryosphere,
11, 2633–2653, 2017

Tuzet, F. et al. (In discussion) :Influence of light absorbing particles on snow spectral irradiance profiles, TCD 2019

Warren S, Brandt R. : Optical constants of ice from the ultraviolet to the microwave: A revised compilation. Journal of Geophysical Research: Atmospheres. 2008