A user (@mmmmansour) has a question about soil data preparation functions.
He launched one simulation using the default load_iSDA_soil_data() function, and the other using load_iSDA_soil_data_alternate().
He obtained contrasted results.
With load_iSDA_soil_data():
With load_iSDA_soil_data_alternate():
To understand these differences, we should explain how these functions are working. Both functions aim at loadig soil physical properties map that are used to obtain a soil water holding capacity. This water holding capacity ("reserve utile", ru) is used in the computations of daily soil water balance.
load_iSDA_soil_data() loads a soil depth data raster (gyga_af_erzd__m_1km.tif, distributed by ISRIC at this address) and a soil texture class (USDA system) raster (iSDA_at_TAMSAT_resolution_zeroclass_1E6.tif), distributed by iSDA at this address. From the USDA soil texture classification, the function then maps soil properties maps (from the file TypeSol_Moy13_HWSD.csv). These soil properties per class include:
"epaisseurProf" : soil depth for deep reservoir
"epaisseurSurf" : soil depth for shallow reservoir
"stockIniProf" : initial water content for deep reservoir
"stockIniSurf" : initial water content for shallow reservoir
"runoff_threshold" : water content theshold from which this soil will have water runoff
"runoff_rate" : rate of water that will be running off
"ru" : reserve utile
That is to say the soil water holding capacity (reserve utile) will be dependant on the soil texture class and the mapping with the csv file.
On the other hand, load_iSDA_soil_data_alternate() will load the reserve utile from gyga_af_agg_erzd_tawcpf23mm__m_1km.tif, which is provided as calculated by ISRIC at this address.
Overall, this means that the water holding capacity of the soil from the second method is provided by a mapped product and will be more continuous in space, whereas the first method will lead to ruptures in the maps, hence some artifacts.
But it is clear that a clearer documentation of these two functions is needed. For the record, the alternate version was quickly implemented for testing purpose, hence this lack of clarity in the documentation.
A user (@mmmmansour) has a question about soil data preparation functions.
He launched one simulation using the default
load_iSDA_soil_data()
function, and the other usingload_iSDA_soil_data_alternate()
.He obtained contrasted results.
With
load_iSDA_soil_data()
:With
load_iSDA_soil_data_alternate()
:To understand these differences, we should explain how these functions are working. Both functions aim at loadig soil physical properties map that are used to obtain a soil water holding capacity. This water holding capacity ("reserve utile", ru) is used in the computations of daily soil water balance.
load_iSDA_soil_data()
loads a soil depth data raster (gyga_af_erzd__m_1km.tif
, distributed by ISRIC at this address) and a soil texture class (USDA system) raster (iSDA_at_TAMSAT_resolution_zeroclass_1E6.tif
), distributed by iSDA at this address. From the USDA soil texture classification, the function then maps soil properties maps (from the fileTypeSol_Moy13_HWSD.csv
). These soil properties per class include:On the other hand,
load_iSDA_soil_data_alternate()
will load the reserve utile fromgyga_af_agg_erzd_tawcpf23mm__m_1km.tif
, which is provided as calculated by ISRIC at this address.Overall, this means that the water holding capacity of the soil from the second method is provided by a mapped product and will be more continuous in space, whereas the first method will lead to ruptures in the maps, hence some artifacts.
But it is clear that a clearer documentation of these two functions is needed. For the record, the alternate version was quickly implemented for testing purpose, hence this lack of clarity in the documentation.