Soil Nutrient Depletin Modelling using Remote Sensig and GIS:
A Case Study in Chonburi, Thailand
4.2. Nutrient Removal by Crops
There were 20 land use and land cover classes derived form Landsat TM 1995. Only 3 main crop areas were included in this study. Usually three to five different sub classes were combined to form one big category. For instance, class no. 2 of paddy, class no. 18 of paddy plus orchard, class no. 24 of second rice, and class no. 25 of paddy plus plantation were merged as paddy. Effective soil depth is. The amount of nutrient up take by crop was identified and taken by the long-term experience of UN/ESCAP, 1993. N, P, and K removal by 3 main crops in kg from 1 ha of soil per year stated in UN/ESCAP, 1993 as follows:
| Paddy (total) |
=N 219 |
P31, |
K258 |
(kg/ha) |
| Cassava(total) |
=N 120 |
P40, |
K187 |
(kg/ha) |
| Sugarcane(total) |
= N45, |
P25, |
K121 |
(kg/ha) |
| Example: Nitrogen uptake by Paddy (kg)= Area extent of Paddy (sq.m)*10-4*219 |
4.3 Nutrient Removal by Soil Erosion
Universal Soil Loss Equation (USLE, Wischmeier and Smith, 1965/1978) was applied in this study as which is by far the most widely used equation. Different approaches to obtain the factors influencing soil loss in USLE and Soil Loss Estimator for Southern Africa(SLAMSA, Elwell 1978a) were reviewed. Of all the approaches only the most suitable and appropriate methods for the factors in USLE were adopted in this study. Due to the lack of available data and information revised USLE (RUSLE, Renard et al., 1991), Water Erosion Prediction Project (WEPP, USDA, 1987), Chemicals, runoff, and erosion from agricultural management systems (CREAMS, Knisel, 1980), and Wind Erosion Prediction Equation (WEPE, Wood ruff and Siddoway 1965; Skidmore and Williams 1991) were omitted. Regarding RUSLE factor like prior land use which requires mass of live and dead roots and buried residues found in the upper 4 in and coefficient representing effectiveness of buried roots and residues information were not available. RUSLE requires a lot of field observation which are normally appropriate for a small watershed area. CREAMS, a combination of process-based and empirically based components, became widely used for field sized areas (Renard, et. al. 1995). In the case of WEPP sediment delivery estimates such as interill erodibility factor K are needed which is normally difficult to manage or not available for large areas. The WEPP is an effort to develop a technology for erosion prediction that extends into the next century (Foster and Land, 1987). WEPP's extensive data requirements and "block box" automated nature may limit its application in developing countries. Its precision and utility are as yet unproven, and its accuracy on steep slopes in the tropics is questionable (David E. Harper).
4.3.1 Calculations of Soil Loss using USLE
4.3.1.1 Rainfall Erosion Index (R )
The following 5 methods were reviewed and used to compute erosivity factor (R ). There were slight variations in all 5 equations.
| From Roose (1975), |
| Mean annual rainfall erosion index (R ) in metric units |
= 0.5 x P x 1.73 (1) |
| From Morgan (1974) |
| Mean annual erosivity (MAE1) (KE>25) (j/m2) |
= 9.28 x P - 8,838 |
| Multiply by I30 (75 mm/h; max. value recommended By Wischmeier) |
= MAE1 x I30 (2) |
| From Foster el. al. (1981), |
| Mean annual EI30 (MAE2) (kg.m.mm)/(m2.h) |
= 0.276 x P x I30 |
| With these units, divide by 100 to give R value in Metric units |
= MAE2/100 (3) |
| From Renard and Freimund, |
| Rainfall erosion index (R ) in metric units |
= 0.0048 x P1.61 (4) |
| From El-Swaify and others 1985, |
| Rainfall erosion index (R ) in metric units |
= 38.5 + 0.35 (P) (5) |
| From Wanapiryarat et. al. 1986, |
| Daily erosivity factor in (y) |
= -3.2353 + 1.789 ln (x) |
| Rainfall erosivity (index (R ) |
= Sy (6) |
Where, P=Mean annual precipitation and x=Daily rainfall data. Results from equation 4 and 5 were found to be about average of all, the first one found to be the highest, and the second and third showed up the lowest. The results from the last equations were discarded and the average of the results from first three equations was taken as R-value in order to minimize the error. Since the rainfall data were not evenly distributed and not may measuring stations were situated in the study area, the calculated values in vector were converted to raster to provide polygon around a set of points.
4.3.1.2 Soil Erodibility Index (K)
K factor is combined effect of processes that regulate rainfall acceptance and the resistance of soil to particle detachment and subsequent transport. Here in this study the K values of different soil series were obtained by the method propose by Ahn, 1978. The approximate values of K were computed by comparing he percentages of clay, sand and silt contents of the soil with the triangular frame of reference provided by the author. The following is the formula developed by Wischmeir and Smith, 1978.
100 K = (2.1 M1.4) 10-4 (12-a) + 3.25 (b-2) + 2.5 (c-3)
where, M = (% silt + % very fine sand) or (100-% clay), a = % of organic matter, b = soil structure class, and c = soil permeability class
4.3.1.3 Topographic Factor (LS)
The combined two factors in USLE the slop length (L) and slope in percent or degree (S) are generally known as topographic factor. Four methods for calculating slope length factor were considered. The first one developed by Wischeier and Smith (1978), the second one developed by McCool, et. al. 1989, Moore and Wilson 1992), the third one by McCool, et. al. 1987), and the last one by SLEMSA, Elwell 1978a).
LS = (x / 22.13)n (0.065 + 0.045 s + 0.0065 s2) (1)
Where, x is the slope length (m) and s is the slope gradient in percent. N = 0.5 for
Slope > 5%, 0.4 for slope 3.5 to 4.5%, 0.3 fro slope l to 3.5%, and 0.2 for slope less than 1%.
LS = (l / 22)m (10.8 Sinb + 0.03) for slopes<9.0%, and
LS = (l /22)m (16.8 Sinb + 0.5) for slopes ³9.0%, (2)
Where, L = slope length in meters,
b = slope gradient in degrees,
M = F / (l + F), and F = (Sin
b / 0.0896) / (3Sin
0.8b + 0.56)
LS = (l / 22.13)0.5, (0.172 s - 0.55) (3)
Where, L = slope length in meters, and s = slope gradient in percent.
LS (X) = (ò l (0.76 + 0.53 s + 0.0765 s2) / 25.65 (4)
Where, l is the slope length (m) and s is the slope steepness in percent. Calculation of combined LS factors based on all those equations showed that the result from the second method gives the best results compared to the others, which give high values for steep slopes.
