Material
Data Used
Inland suitability evaluation, adequate and quality data is extremely important to supply a highly precious land suitability map of the study area. Thus, relevant data including soil physical characteristics, topographic, accessibility, and water source were collected (Table 2). These data were collected from the FAO, and Earthexplorer.org (Table 1). Eight input parameters, which were used for analysis, were selected based on data availability, agricultural history, literature review, and native condition of the study area (Table 1). These are slope, elevation, soil depth, soil drainage, soil type, soil texture, distance from the water source, and land use /land cover. The physical parameters were extracted from the food and agricultural organization (FAO) includes soil depth, soil drainage, soil texture, and soil type. Topographic condition is one among the factors, which will affect the suitability of land for agricultural activity. Slope can control the speed of erosion and deposition whereas elevation can control the climatic condition within the region. The slope gradient and elevation are sorts of topographic parameters, which were extracted from the SRTM DEM (30m resolution). Distance from the river was generated using Euclidian distance. The land use and land cover map of the study area were prepared from Landsat 8 satellite images using ERDAS and GIS 10.3.
Table 1 Data used
Data type
|
Criteria
|
Data sources
|
Data format
|
Soil physical parameter
|
Soil texture
Soil depth
Soil type
Soil drainage
|
http://www.fao.org/soils-portal/data-hub/soil-maps-and-databases/faounesco-soil-map-of-the-world/en/
|
vector
|
Topography
|
Slope
|
USGS
|
raster
|
Elevation
|
USGS
|
raster
|
Landsat 8 satellite image
|
Land use land cover
|
USGS
|
raster
|
Water
|
Distance to river
|
ethiogis-mapserver.org/
|
vector
|
Methods
For this research, data collection, satellite image analysis, land suitability factor evaluation, and mapping, and GIS-based analytical hierarchy process land suitability modeling were applied. In GIS-based land suitability analysis, the input file should be converted into an identical projection and same pixel sizes. For this purpose, after collecting the info, all the parameters were converted into a 30m resolution raster with an equivalent projection type. The rasterized parameters were reclassified and standardized supported the importance of every sub-criteria. After all, the suitability of the parameters was reclassified and rescaled which range from 1-7 (Effat and Hassan, 2013). The size of 1 refers to rock bottom importance, but 7 is that the higher importance. Besides, projection and pixel size, geo-database is extremely essential in data management within the GIS environment, therefore, two geo-database were built. After the database was built, an evaluation of the connection between land suitability and factors also because the significance of every parameter and sub-criteria was performed using the principle of the analytical hierarchy process. Therefore, eight land suitability factors were reclassified and rescaled. Then, the load of every criterion was calculated employing a pairwise comparison matrix. The weighted parameters (criteria) were sum up using weighted overlay methods in ArcGIS to get a land suitability index. After all, the land suitability index map was reclassified into four suitability classes using the natural break method within the ArcGIS tool. These are highly suitable (S1), suitable (S2), marginally suitable (S3), and unsuitable class (S4). The general methodology flow chart for this study is summarized in Fig.6.
Evaluation of Criteria Maps
Topographic parameters (slope and elevation)
The map of elevation and slope was generated from the Shuttle Radar Topography Mission (SRTM) digital elevation model (DEM) data of 30 m resolution which is downloaded from Earth explorer.org (Moore and Hansen 2011; Gorelick, 2013). Considering FAO (2003) manual, the slope and elevation of the study area were classified into different suitability classes (Table 3). The slope is one among the many parameters in land suitability analysis for surface irrigation which may also affect the irrigation land preparation, irrigation operation, production costs, method of irrigation, soil depth, and erosion (USDIBR, 2003; Kassaye et al., 2019). within the present study, the slope is taken into account as input parameters that were extracted from 30 m DEM resolution which was classified into four suitable classes of 0-2%, 2-5%, 5-8%, and >8% (USDIBR, 2003; Kassaye et al., 2019). The slope angle can affect the suitability of the land for surface irrigation in terms of land preparation for irrigation and operation also as moisture content and erosion rate. The steeper the slope, the harder surface irrigation operation and therefore the higher erosion rate. This results in the less agricultural productivity of the farm field.
Soil
As stated by Dagnenet (2013); USDIBR (2003), the soil is a crucial factor to work out the suitability of land for sustainable surface irrigation and agriculture. The soil can affect productivity capacity, production and irrigation operation, and development (Kasaye et al., 2019). The physical soil data were downloaded from Harmonized World Soil Database (HWSD). The physical characteristics of soil are often evaluated to work out soil suitability of irrigation (USDIBR 2003). The consequences of soil for irrigation and agriculture are primarily associated with soil fertility, toxicity, moisture content, depth, and credibility (USDIBR, 2003). The first factors are soil-moisture relationships, toxicity, fertility, depth to gravel and cobble, continuing layer, and erosion hazard. Therefore, for this study, physical (soil type, soil drainage, soil depth, and texture classes) are primary soil factors that extracted from HWSD (Table 3). the detail of degree of significance of parameter classes are summarized in Table 1 that has been organized by International Institute for Applied analysis (IIASA), ISRIC World Soil Information Institute of Soil Science Chinese Academy of Sciences (ISSCAS), Joint research facility of European Commission (JRC), and FAO (Nachtergaeele et al., 2009, Kassaye et al., 2019).
Land use land cover
Land use is one among the important parameters in land suitability analysis for surface irrigation. The land use map was generated from Landsat 8 satellite image that downloaded from the USGS website. It had been classified into four suitable classes of highly suitable (S1), suitable (S2), marginally suitable (S3), and not suitable (S4). Cropland was grouped as highly suitable (S1), but grassland is assessed as moderately suitable (S2).Woodland/Shrub/bush was classified as marginally suitable (S3) because it required initial investment for land preparation (Kassaye et al., 2019). The Forest/Barren land/Water body/Settlements are classified as not suitable (S4).
Distance to Streams
Evaluation and determination of the space to the water source are one among the foremost important inland suitability analyses for surface irrigation. Distance from water source was generated using Euclidean distance in ArcGIS which was classified into 0-1.23 (S1), 1.23-2.5 (S2), 2.5-3.7 km (S3), and > 3.7 (S4).
Table 2 Significance of physical factors for surface irrigation (Kassaye et al., 2019)
Parameters
|
Parameters classes
|
Degree of significance
|
Sources
|
Drainage
|
Well
Moderately Well
Imperfectly
Poor
excessive
|
Optimum
moderate marginal
Not ideal for upland crops
has little significance
|
Nachtergaele et al., 2009
Kassaye et al. 2019
|
Soil Depth(cm)
|
< 10
.10 - 50
50 - 100
> 100
|
very low
low
Marginal
Optimum
|
Mandal et al., 2017
Kassaye et al. 2019
|
Soil Type
|
Chromic Luvisols
Humic Nitosols Eutric Vertisols Haplic Calcisols Rendzic Leptosols
Lithic Leptosols
|
Optimum
Moderate
marginal low
very low
very low
|
FAO and
UNESCO,1988
Kassaye et al. 2019
|
Texture Classes
|
Sandy Loam
Loamy
Sandy Loam
Silt Loam
|
Marginal
low
Optimum
Moderate
|
USDIBR, 2003
Mandal et al., 2017
Kassaye et al. 2019
|
Euclidean Distance (km)
|
0 - 5
5 - 10
10 - 20
> 20
|
optimum
moderate
marginal
low
|
Mandal et al, 2017
Kassaye et al. 2019
|
LULC
|
Rangelands
Farmland
Dispersed Forest
Settlement
Bush
Bare land
|
Moderate
Optimum
Low
Not optimal
Marginal
Not optimal
|
Mandal et al, 2017
Kassaye et al. 2019
|
Slope (%)
|
0 - 2
2 - 5
5 – 8
>8
|
Optimum
Moderate
Marginal
Low
|
Mandal et al, 2017
USDIBR, 2003
Buhari, 2014
Kassaye et al. 2019
|
Parameter Reclassification
Reclassification will tell us the degree of suitability of every class by partitioning it into different parts supported evidence. Parameter reclassification usually needs modeling experience and adequate field knowledge (Forkuo, 2011). Besides, assistance from other experts like from agricultural exports, and climatologist is extremely important to urge critical information during which soil texture, depth, type, drainage, elevation, slope, distance to a water source, and land use conditions are favorable for surface irrigation land suitability mapping. The values within the parameters must be prioritized in a single raster. Values in a particular raster might not be fit or fit the proposed objective (ESRI, 2017). For instance, the soil texture class (clay) is grouped under a highly suitable soil class (S1) compared to other classes. During this study, the parameter reclassification was performed using the size of 0.5 to 7 using the rule of land classification of ICARDA (2019) in Table 2. As shown in Table 3, all criteria are reclassified to different land suitability categories considering their suitability for surface irrigation. the number seven is assigned for highly suitable land class (S1), six is assigned for suitable land class (S2), three is assigned for marginally suitable land (S3) one is assigned for unsuitable land class at current condition (S4) and 0.5 is assigned for permanently unsuitable land class (S5).
Table 3 Reclassified parameters
Parameters
|
Drainage Class
|
Degree of importance
|
Land suitability
|
Soil drainage
|
Excessive (LP)
|
0.5
|
S5
|
Poor (VR)
|
1
|
S4
|
Imperfect (CM)
|
3
|
S3
|
Moderate (NT)
|
6
|
S2
|
Well (LV)
|
7
|
S1
|
Parameter
|
Soil type Class
|
Degree of importance
|
Land suitability
|
Soil type
|
LP
|
1
|
S4
|
VR
|
3
|
S3
|
NT
|
6
|
S2
|
CM/LV
|
7
|
S1
|
Parameter
|
Elevation Class
|
Degree of importance
|
Land suitability
|
Elevation
|
2225-2821
|
1
|
S4
|
1862-2225
|
3
|
S3
|
1609-1862
|
6
|
S2
|
1168-1609
|
7
|
S1
|
Parameter
|
Distance to stream
|
Degree of importance
|
Land suitability
|
Distance to stream
|
<1.23
|
7
|
S1
|
1.23-2.5
|
6
|
S2
|
2.5-3.7
|
3
|
S3
|
> 3.7
|
1
|
S4
|
Parameter
|
Texture
|
Degree of importance
|
Land suitability
|
Soil texture
|
Sandy loam
|
3
|
S3
|
Loam
|
6
|
S2
|
Clay
|
7
|
S1
|
Parameter
|
Depth
|
Degree of importance
|
Land suitability
|
Soil depth (cm)
|
10
|
3
|
S3
|
100
|
6
|
S2
|
Parameter
|
LULC
|
Degree of importance
|
Land suitability
|
LULC
|
Forest/Barren land/Water body/Settlements
|
1
|
S4
|
Woodland/Shrub/bush
|
3
|
S3
|
Grassland
|
6
|
S2
|
Cropland
|
7
|
S1
|
Parameter
|
Slope
|
Degree of importance
|
Land suitability
|
Slope (%)
|
<2
|
7
|
S1
|
2-5
|
6
|
S2
|
5-8
|
3
|
S3
|
>8
|
1
|
S4
|
Land Suitability Classification Structure
As stated by FAO (1976), Land suitability is that the fitness of a given sort of land for an outlined use. FAO (1976 and 2007) developed land suitability classes of highly suitable to not suitable lands (Table). These classes are generally classified into suitable and not suitable (S4 or N). In 1976 and 1983, FAO further classified the land into three and two classes respectively. This classification was made considering the advantages and limitations (Table 4). These are highly suitable (S1), moderately suitable (S2), marginally suitable (S3), temporarily not suitable (S4/N1), and permanently not suitable (S5/N2). Source: FAO, 1976 and 1981.
Table 4 Land suitability classification structures (FAO, 1976 and 1981)
Class
|
Description
|
S1 Highly Suitable
|
Land without significant limitations. This land is the best possible and does not reduce productivity or require increased inputs.
|
S2 Moderately Suitable
|
Land that is clearly suitable but has limitations that either reduce productivity or require an increase of inputs to sustain productivity compared with those needed on S1 land
|
S3 Marginally Suitable
|
Land with limitations so severe that benefits are reduced, and/or the inputs required sustaining production need to be increased so that this cost is only marginally justified.
|
N1 Currently Not Suitable
|
Land having limitations which may be surmountable in time, but which cannot be corrected with existing knowledge at currently acceptable cost; the limitations are so severe as to preclude successful sustained use of the land in the given manner
|
N2 Permanently Not Suitable:
|
Land having limitations that appear as severe as to preclude any possibilities of successful sustained use of the land in the given manner.
|
Preparation of Datasets or Criteria Maps
Parameter preparation is one among the important steps of inland suitability analysis in surface irrigation. In this study, one database was built for criteria maps, which contains eight parameters. The parameters were converted into an equivalent pixel size and projection system. This study considers eight parameters like slope (Figure 3d), elevation (Figure 2b), land use (Figure 3c), soil texture (Figure 5h), depth, type (Figure 4f), and drainage (Figure 4e). The physical soil maps with attributes were extracted from HWSD and converted into roasters with an equivalent pixel size and projection using GIS. The slope, elevation, and distance from the water source are generated from 30 m DEM resolution. The land use /land cover map was derived from Landsat 8 satellite image (Figure 3c).
Analytical Hierarchy Process (AHP) and Pairwise Comparison Matrix
One of the semi-quantitative methods that won’t to determine land suitability for surface irrigation is named AHP. AHP may be a structured tool that helps to research difficult decisions supported mathematics and psychology (Cho et al. 2015; Nguyen et al. 2015; Saaty 2000). during this study, a pairwise comparison matrix was conducted using excel to supply weights for every parameter or criteria considering Saaty’s ranking scale (Luu et al. 2018; Saaty 2008). Before a pairwise comparison matrix is performed, scale is extremely important. For this, Saaty (2008) was developed a scale, which ranges from 1 to 9. This enables for the equitable comparison of the intensity of land suitability. One indicated equal importance of parameters whereas nine indicated extreme importance of the factor or the parameters (Malczewski, 1999; Saaty, 1980). During this study, an 8x8 comparison matrix was performed by assigning a worth from the range of 1-9 by comparing the parameter within the row to the parameter within the column (Table 5). for instance , from the matrix (Table 4), 9 was assigned for distance to water source compared to texture, depth, type, drainage, and LULC while 1/9 was assigned for texture, depth, type, drainage, and LULC compared to the space to a water source. After a completed comparison among all possible criteria or parameters, normalization was performed by dividing the cell by the column total (Saaty, 1977). Normalization of the matrix is extremely important to work out the standards weight. The load of every criterion or parameter is extremely important to gauge the importance of every parameter ashore suitability analysis or evaluation. This will be calculated using AHP calculators during a pairwise comparison matrix (Saaty, 1977). The load of every parameter is employed to gauge the relative importance of every criterion. Within the current study, Saaty’s method of eigenvectors or relative weights is calculated from the normalization table (Table 6). Within the analytical hierarchy process, consistency of calculated weight for every parameter or criteria is one among the issues that require examination (Luu et al. 2018; Saaty 2001). Therefore, after the load of the criterion was calculated, the consistency of the calculation was evaluated to gauge the strength of the tactic (Saaty, 1990; Garcı ́a et al., 2014). This will be determined using consistency ratio or relationship (CR) which may be estimated using Eq. 1.
(1)
The CI was calculated using Eq. 2.
CI is that the consistency index (0.1396) and RI is that the random index; n may be a number of parameters (n=8) and λmax (8.977) is that the principal Eigenvalue. RI (1.41) is that the average of the consistency index during which it depends on the matrix given by Saaty (1977). CR is that the consistency ratio, which refers to the degree of the consistency or inconsistency of the parameters or criteria (Chen et al., 2010b). When CR>0.1, the pairwise comparison matrix is not correct and it needs revision. Supported the reason, the CR value for the present study area is <0.1 which is CR= 0.098987, therefore the matrix is correct. Then after evaluated the consistency condition of the weighted criteria, the land suitability class was generated using weight sum methods in ArcGIS (Rahmati et al., 2016c) as showed in Eq.3.
(3)
Where LSM is land suitability map, n is number of criteria or parameter, Wi is criteria weight, and Xi is criteria or parameter.
Table 5 Comparison matrix
|
Distance to Water source
|
Slope
|
Elevation
|
Soil texture
|
Soil drainage
|
Soil depth
|
LULC
|
Soil type
|
Distance to Water source
|
1
|
5
|
7
|
9
|
9
|
9
|
9
|
9
|
Slope
|
0.2
|
1
|
3
|
2
|
5
|
7
|
7
|
7
|
Elevation
|
0.14
|
0.33
|
1
|
2
|
7
|
5
|
5
|
5
|
Soil texture
|
0.11
|
0.5
|
0.5
|
1
|
5
|
7
|
7
|
7
|
Soil drainage
|
0.11
|
0.2
|
0.14
|
0.2
|
1
|
1
|
2
|
5
|
Soil depth
|
0.11
|
0.14
|
0.2
|
0.14
|
1
|
1
|
2
|
2
|
LULC
|
0.11
|
0.14
|
0.2
|
0.14
|
0.5
|
0.5
|
1
|
2
|
Soil type
|
0.11
|
0.14
|
0.2
|
0.14
|
0.2
|
0.5
|
0.5
|
1
|
Total
|
1.89
|
7.45
|
12.24
|
14.62
|
28.7
|
31
|
33.5
|
38
|
Number of comparisons = 28
|
|
|
|
|
|
|
Consistency Ratio CR = 9.9%
|
|
|
|
|
|
|
Principal Eigen value = 8.977
|
|
|
|
|
|
|
Table 6 Normalized table
|
Distance to water source
|
Slope
|
Elevation
|
texture
|
drainage
|
depth
|
LULC
|
type
|
Eigenvector of Weights
|
Weight %
|
Distance to water source
|
0.53
|
0.67
|
0.57
|
0.62
|
0.31
|
0.29
|
0.27
|
0.24
|
0.44
|
44%
|
Slope
|
0.11
|
0.13
|
0.25
|
0.14
|
0.17
|
0.23
|
0.21
|
0.18
|
0.18
|
18%
|
Elevation
|
0.07
|
0.04
|
0.08
|
0.14
|
0.24
|
0.16
|
0.15
|
0.13
|
0.13
|
13%
|
texture
|
0.06
|
0.07
|
0.04
|
0.07
|
0.17
|
0.23
|
0.21
|
0.18
|
0.13
|
13%
|
drainage
|
0.06
|
0.03
|
0.01
|
0.01
|
0.03
|
0.03
|
0.06
|
0.13
|
0.05
|
5%
|
depth
|
0.06
|
0.02
|
0.02
|
0.01
|
0.03
|
0.03
|
0.06
|
0.05
|
0.04
|
4%
|
LULC
|
0.06
|
0.02
|
0.02
|
0.01
|
0.02
|
0.02
|
0.03
|
0.05
|
0.03
|
3%
|
Soil type
|
0.06
|
0.02
|
0.02
|
0.01
|
0.01
|
0.02
|
0.01
|
0.03
|
0.02
|
2%
|