`mrsofc` mass content calculation from VMR

[sol1105]

For ICON-CLM, the variable mrsofc (Capacity of Soil to Store Water (Field Capacity), soil_moisture_content_at_field_capacity) is output as volume mixing ratio, i.e. with units m3/m3.
We have now some concerns / questions on how to convert it to a mass content, as requested.

The formula would be:
mrsofc = soil_field_cap_vmr * soil_depth * 1000 kg/m3

The CF conventions state:

soil_moisture_content_at_field_capacity (canonical unit: "kg m-2")
"moisture" means water in all phases contained in soil. "Content" indicates a quantity per unit area. The "soil content" of a quantity refers to the vertical integral from the surface down to the bottom of the soil model. For the content between specified levels in the soil, standard names including content_of_soil_layer are used. The field capacity of soil is the maximum content of water it can retain against gravitational drainage.

So we wonder if we should simply use the bottom of the soil model as soil_depth. This would however lead to quite high values compared to what was submitted by modeling groups for past CMIP and CORDEX phases.

For the MPI-ESM soil model JSBACH in CMIP6, as soil_depth, the depth_to_bedrock was used for this calculation. For ICON-CLM however, we do not have a depth_to_bedrock available. So, we could:

• use the root_depth instead
• use a fixed depth of eg. 1m
• use the full vertical extent of the soil model (15.34m in our case), as CF suggests
• (not provide the variable)

What option do you suggest, or do you have an alternative suggestion we did not yet think about?

Thank you and kind regards,
Martin

[sol1105]

@gnikulin Is there any preference to eg. use the root_depth? Else we will not be able to provide this variable.

[sol1105]

If there are no objections we would use the depth of the lowest hydrologically active layer for the calculation. We do not know if it will allow meaningful intercomparison, but we will add a variable comment regarding the calculation.

[larsbuntemeyer]

I have the same questions for REMO actually (also dimensioless output). I once asked copilot about this:

---

To convert the field capacity of soil from a volumetric water content (usually expressed as a percentage) to kilograms per square meter (kg/m²), you need to know the bulk density of the soil. The bulk density is the mass of soil per unit volume, typically expressed in grams per cubic centimeter (g/cm³) or kilograms per cubic meter (kg/m³).

Steps to Convert Field Capacity to kg/m²

• Determine the Field Capacity (FC): This is usually given as a percentage (e.g., 25%).
• Determine the Bulk Density (BD): This is typically given in g/cm³ or kg/m³.

Calculate the Water Content in kg/m²:

• Convert the field capacity percentage to a decimal (e.g., 25% becomes 0.25).
• Multiply the field capacity by the bulk density to get the water content per unit volume.
• Multiply by the depth of the soil layer (in meters) to get the water content per unit area.

---

So i came up with the same factor of 1000 kg/m3 but not knowing what depth to use. I can try to stick to your approach and maybe compare. For now, we decided not to cmor this (didn't have this in CMIP5 either) but maybe it makes sense if i use the same approach.

[jesusff]

We have similar discussions with WRF (CORDEX-WRF-community/fps-urb-rcc#11). I think it is key to consider if you have a volumetric mixing ratio or a volumetric fraction. Both are dimensionless (m3/m3), but I don't see the need for the soil density in the second case, as it refers (does it?) to a fraction of the total soil column volume.

[gnikulin]

There were several requests to provide mrsofc and it was simply copy/paste from CMIP. I would use soil_depth or depth_to_bedrock to calculate mrsofc, not sure about root_depth , not an expert in this area. The 3D variable mrsolmust provide "The mass of water in all phases in each soil layer (3D variable)". Can the same soil layers be used ?