Particle-Size Distribution Models for the Conversion of Chinese Data to FAO/USDA System

Table 4

Number of cases as the best model with the smallest AIC value for each soil textural class in T2 (a) and T3 (b) schemes†.

(a)

Texture†	AD	F3P	F4P	ML	ONL	ORL	S	VG	VGM	W	SELF	Sum
SL	1§	1§	0	0	0	0	0	0	0	0	0	2
HL	17	5	7	4	8	2	32§	5	5	14	0	99
ML	8§	4	1	4	1	6	3	0	0	1	0	28
LL	2§	1	0	0	0	1	1	1	0	0	0	6
HC	38§	6	19	1	1	0	4	0	0	3	8	80
MC	62§	8	50	5	9	1	31	3	1	10	5	185
LC	79§	0	60	7	12	2	67	8	6	29	1	271
Sum	207§	25	137	21	31	12	138	17	12	57	14	671

(b)

Texture‡	AD	F3P	F4P	ML	ONL	ORL	S	VG	VGM	W	SELF	Sum
TS	2§	1	1	0	0	0	0	0	0	0	0	4
LS	0	1	2§	0	0	0	0	0	0	0	0	3
SL	9§	2	1	0	1	2	5	0	0	0	0	20
HL	74§	1	16	10	22	7	45	0	3	23	6	207
ML	53§	7	9	4	6	15	14	0	1	13	1	123
LL	21§	4	7	1	0	5	4	1	0	4	0	47
HC	40§	7	14	0	0	0	12	0	0	7	11	91
MC	90§	1	37	2	9	7	44	0	0	30	9	229
LC	143§	5	39	0	39	3	85	1	4	39	8	366
Sum	432§	29	126	17	77	39	209	2	8	116	35	1090

†For example, if the best model according to AIC value of a soil sample is the AD model, the number of the AD model in the table will increase by one. If a model has a big number in a soil textural class, it has advantages in fitting the PSD curve compared to models with small number.
‡Textural classes of Katschinski system: TS = tight sand, LS = loose sand, SL = sandy loam, HL = heavy loam, ML = moderate loam, LL = light loam, HC = heavy clay, MC = moderate clay, and LC = light clay.
§The biggest number of cases as the best model for the specific soil texture class.