Lack of understanding about the use of soil data in SARRA-Py

[Nenwala]

Hello,
I'm having some difficulty understanding how soil data is used in the SARRA-py model.
Taking the example of the notebook "Exemple 2 : recalibrating crop simulation parameters based on optimization - North Cameroon Maize 1999-2022", I noticed that the soil file used is: "file_paramTypeSol = "USA_iowa_V42.yaml"", as defined in section 4. Preparing and launching simulations.
Looking into the concerned file (in the params/soil directory), I see it contains parameters such as "epaisseurProf", with numerical values.
Could you please clarify what these values represent?
Are they averages? And if so, over which geographic areas were these averages computed?
Thank you.

[codename5281]

Hello @Nenwala ,

Thanks for reaching out,

Let me clarify by explaining the other way round and exposing how soil parameters are loaded in the notebook you work with, that is indeed based on the Exemple 2 notebook.

In this notebook, soil parameters are initialised and loaded using the following line :
base_data = load_iSDA_soil_data_alternate(base_data, grid_width, grid_height) # load soil parameters

This function does the following things:

• loads information about depth of the reserve utile (from the Africa SoilGrids Root zone depth data, stored in the file ./data/assets/gyga_af_erzd__m_1km.tif) in the profRu variable
• initializes the Soil surface reservoir depth (mm) (epaisseurSurf variable) with a standard 200mm value
• initializes the Initial water stock in the deep surface reservoir (mm) (stockIniProf variable) as 0 mm
• loads soil texture classes identifyer (from the iSDA Africa soil classes data, stored in the file ./data/assets/iSDA_at_TAMSAT_resolution_zeroclass_1E6.tif), and associates runoff threshold (SeuilRuiss) and runoff rate (PourcRuiss) properties based on a correspondance table (./data/assets/TypeSol_Moy13_HWSD.csv) coming from legacy SARRA-H models
• plant available water content (RZPAWC) is loaded from ./data/assets/gyga_af_agg_erzd_tawcpf23mm__m_1km.tif
• finally, reserve utile ru is calculated by dividing RZPAWC by depth of reserve utile

In this notebook, as all data are loaded from map datasets, none of the soil parameters loaded by the load_YAML_parameters function are used when defining soil properties.

In conclusion, these can be disregarded.

As for the epaisseurProf variable, it is not used in the calculations anymore.

As you see, there is still a lot of cleanup to do in this code, given that variables were renamed at some point in the development.

Hoping i answered to your questions,
Best,
J.

[Nenwala]

Bonjour @codename5281
Merci pour cette explication complète et explicite qui répond bien à ma demande.

Nenwala