Methodology

Concept

The Food Agriculture Organization (FAO) sets standards on the requirements of the crops for cultivation. The first five crops available in the list are the following:

1. Alfalfa
2. Avocado
3. Bamboo
4. Banana
5. Barleys

In order to tell if a given land unit is suitable for a particular crop, the characteristics of the said land are compared to the crop requirements, to see the quality of the land base on the suitability scores. As for illustration, let us consider the popular crop, Soya, with the following Terrain characteristics.

Table 1. Soya Terrain Characteristics.

Code	S3	S2	S1	S1	S2	S3	Weight class
Slope1	6	4	2				3
Slope2	16	8	4				2
Slope3	30	16	8				3
Flood	2.5	2	1				1
Drainage4	4	3	2				2
Drainage5	4	1	2				2
SlopeD	4	3	2				2

And we are going to test the following land unit's characteristics,

Table 2. Sample Land Unit Land Terrain Characterictics.

LU	SlopeD	CFragm	SoilDpt	SoilTe	CECc	SumBCs	pHH2	BS	OC	Flood
1	1	0	45	9	18.6	3.3	5.4	50	1.4	3

In the above table, LU stands for Land Unit and the proceeding columns are the available characteristics of the land. Hence, for first LU we have two factors available in Soya requirements, namely i. SlopeD and ii. Flood. Practically speaking, evaluating the suitability class of the first Land Unit can be done in just a glance. For SlopeD, for example, base on the methodology of LUSET, the three suitability class (S1, S2 and S3) for SlopeD have the following domains,

Table 3. SlopeD Factor Domain using LUSET Method.

Class	Domain
S1	< 2
S2	2 - 4
S3	4 - 6

In that case, if LU1 has SlopeD = 1, then it has S1 suitability class, since 1 is within the domain of S1. On the other hand, for Flood, again using LUSET we have the following domain,

Table 4. Flood Factor Domain using LUSET Method.

Class	Domain
S1	< 1
S2	1 - 2
S3	2 - 2.5

In this case, if LU1 has Flood = 3, then it is outside the domain of any suitability class above. Implies not suitable. Because domains on every factor are different, thus it is tiresome to do the assignment from factor to factor for every land units. A remedy to this is to generalize the assignment by implementing suitability scores which can be divided into different domain of suitability class. For LUSET, the span of the suitability scores are defined below,

Table 5. Suitability Scores Domains on LUSET.

Class	Suitability Scores	Description
S1	[85% - 100%]	Highly Suitable
S2	[60% - 85%)	Moderately Suitable
S3	[40% - 60%)	Marginally Suitable
N	[0% - 40%)	Not Suitable

In ALUES, however, the fuzzy membership function for standard assignment of scores are utilized. For example, using Triangular membership function, we can range the scores in percentage into the following:

Table 6. Suitability Scores Domains on ALUES.

Class	Suitability Scores	Description
S1	[75% - 100%]	Highly Suitable
S2	[50% - 75%)	Moderately Suitable
S3	[25% - 50%)	Marginally Suitable
N	[0% - 25%)	Not Suitable

Notice the difference with that in LUSET, here we have assigned the domains into equal parts, by 25%. And below is the equivalent fuzzy model diagram,

For SlopeD we have the following diagram,

Now by default, the formulas for min and max are the following:

min is set to zero since from our investigation in the CropInfo.xls file of FAO, most of the units of the factors are in cm, mm, %, degree Celsius, and the likes, which of course excludes negative values, although these are possible for temperature, such as in cold condition (below zero), but it is doubtful to think for a crop that would require that condition. ALUES aims to make everything as general as possible, so the min value can be modified by the user, either uniform min value for all factors or different min value for every factor. This ability is not yet available in the landSuit function, but that is our vision. As for max, ALUES aims for a non-bias formula, thus we simply take the difference between a and b, and b and c, to obtain the distance of every interval. Then we divide this by two to extract the average distance. Finally, adding this to c gives us the maximum value.

Again, ALUES values generality in the function as much as possible. So users can also modify the domains of the default assignment of suitability scores. For example, users can mimic the method of LUSET by adjusting the four intervals of suitability class in Table 6 to that in Table 5.

Note: Some features of ALUES mentioned here are still under development. This page is under development so it may change any time.