Yuki sulpher #26
Yuki sulpher #26
Conversation
The fix consists of averaging using the soil type and ignoring region
For arable an sbal larger than 0 should always give an ind of 1 because no fertilization is needed, because the soil delivers enough S
|
Combinations of soil type and region that do not exists in table 6.2 of Handboek Bodem & Bemesting are now solved by using the average value of the soil type and ignoring the region |
|
The evaluation of I_C_S for arable should not be changed. An |
|
@yuki-nmi shall we merge this PR? |
|
I am still doubting to use the abrupt shift of I_C_S from 0 to 1 for arable land. Theoretically it's true that S availability is enough whtn sbal > 0. However, sbal is derived from a continuous variable SLV and several other constants from lookup table, and a slight change in SLV can results in either I_C_S =1 or I_C_S = 0. This sounds like an artifact. |
|
I changed the evaluation funtion to a logistic one to get a more smooth transition |
Some B_LG_CBS x B_B_AK combinatoins miss 'clust' values (which determines 'benodigde S uit grond en bemesting (‘slv_av’)'. This results NA values for I_C_S. Clust numbers need to be assigned to all combinations.