no atmospheric snow (mm/s) but still snowcover, why?

[Tanja]

I am running CLM in single-point mode. In this setup with C3 grass.
When setting precip_repartition_nonglc_all_rain_t and precip_repartition_nonglc_all_snow_t to 0ᵒC and -2ᵒC, respectively, the simulated snow cover (FSNO) is considerable predominant compared to what we observe at the site. To study the dynamics when there is no snowfall (SNOW), I lowered precip_repartition_nonglc_all_rain_t and precip_repartition_nonglc_all_snow_t to -20ᵒC and -22ᵒC. I was surprised to observe that CLM correctly did not simulated snowfall (mm/s), however still simulate snow cover (FSNO) despite no snowfall over a six years period. Can somebody explain me how this can be, and if I can avoid snowcover?

 

[oleson]

That is odd. Is your simulation initialized with a lot of snow by chance?

 

[Tanja]

There is snow in the beginning of the period, but as can be seen at the figures, the snowcover is not continuous through out the simulation period. The only difference between the two model runs shown in the figures are the change of precip_repartition_nonglc_all_rain_t and precip_repartition_nonglc_all_snow_t as described above.

I suppose that the LAI=0 is because of the snowcover, and this LAI=0 was what I wanted to avoid. Any idea?

Figure 1: precip_repartition_nonglc_all_rain_t=0 ᵒC, precip_repartition_nonglc_all_snow_ t=-2 ᵒC



Figure 2: precip_repartition_nonglc_all_rain_t=-20 ᵒC, precip_repartition_nonglc_all_snow_ t=-22 ᵒC

 

[swensosc]

Can you plot H2OSNO? There are conversions that occur with surface water, as well as frost deposition.

 

[Tanja]

Here I plotted H2OSNO.

In another test (figure 5) I also tried to set h2osno_max = 0.01. This resulted in H2OSNO>0.01 mm. I find it very odd.



Figure 3: precip_repartition_nonglc_all_rain_t=0 ᵒC, precip_repartition_nonglc_all_snow_ t=-2 ᵒC


Figure 4: precip_repartition_nonglc_all_rain_t=-20 ᵒC, precip_repartition_nonglc_all_snow_ t=-22 ᵒC



Figure 4: precip_repartition_nonglc_all_rain_t=0 ᵒC, precip_repartition_nonglc_all_snow_ t=-2 ᵒC, h2osno_max=0.01

 

[swensosc]

Can you plot H2OSFC to see whether surface water is freezing and being incorporated into the snowpack?

 

[Tanja]

Is there a way where I can avoid that freezing surface water are being incorporated into the snowpack?

Figure 5: precip_repartition_nonglc_all_rain_t=-20 ᵒC, precip_repartition_nonglc_all_snow_ t=-22 ᵒC

 

[swensosc]

One way might be to set the microtopography parameter to zero. That should reduce the storage capacity for surface water to zero, and it should go to runoff. In src/main/initVerticalMod.F90, you could set col%micro_sigma(c) to zero and see if that eliminates the surface water.