Description of the Research Area
The study areas are in Ethiopia's Amhara regions, in the west Gojjam and Awi districts. The regions are located in a latitude of 10° 10'-11° 0'North and a longitude of 36° 45'-37° 25'East, 122 kilometers south of Bahir Dar and 447 kilometers north of Addis Abeba. This watershed included a total area of 1057.897 km2, with altitudes ranging from 2939 meters to 801 meters above mean sea level.
The watershed is divided into five agro climatic zones: moist lower Kolla, moist upper Kolla, moist Weyena Dega, wet Weyena Dega, and wet Dega. 0.05 percent moist lower kolla, 5.38 percent moist upper kolla, 63.92 percent moist Weyena Dega, 3.12 percent Wet Dega, and 27.51 percent wet Weyena Dega of the research areas are covered by the watershed. The studied areas' rainfall patterns are unimodal, and the primary wet seasons are summer (June to September). The research area receives 1951mm of rain per year on average. The area's yearly average lowest and maximum air temperatures are 9.2°C and 25.1°C, respectively.
Data used
As demonstrated in Table 1.1, various data were employed for different determinations in this inquiry.
Table 1
No
|
Data Used
|
Purpose
|
Source
|
1
|
Soil data
|
Land suitability
|
Amhara soil map
|
2
|
Land use land cover
|
Land suitability
|
Landsat8satelliteImage, 2019year
|
3
|
Stream flow data
|
Estimate water resource potential
|
more
|
4
|
DEM
|
Delineate watershed and generate slope
|
Amhara DEM map
|
5
|
Metrological data
|
Estimate irrigation water requirement
|
Metrological Agency
|
The software and tools that were employed
The software applications that are used;
Arc GIS was used to map the potential irrigable region, HBV was used to show available water, CROPWAT 8 was used to compute crop water requirements, and RStudio was used to estimate missing data.
Methodology
Identifying places that could be irrigated
The determination of suitable for irrigations was done elsewhere, taking into account soils, land use land covers, slops, and distances between irrigated regions with potential capacity areas as factors. Individual appropriateness of each element was first assessed, then weighted to determine possible irrigable locations.
Slope suitability analyses
The slopes are the most important topographical criteria in determining whether or not a piece of land is suitable for irrigation. The DEM of the research areas was trimmed as of Amhara DEM with a 30*30 meters resolution by masking layer of watershed boundary to create slopes appropriateness maps. The spatial analysis slope tools in ArcGIS produced a slopes map of the watersheds as a consequence.
Soil suitability analysis
The most important feature of a land's suitability for surface irrigation development is the soil. Depth, drainage, and texture classes were the most important significant restrictions of the soil throughout the watershed, and they were all used extensively in the suitability analysis. With the FAO land and water bulletin, soil suitability rates were calculated based on the FAO technique for land estimation (Meaden, 1991). (Reynolds et al., 2000). Table 2.4: Rating of soil suitability factors (FAO, 1991)
Land use land cover
Land usage and land cover of the areas are also factors that have been used to determine land suitability for irrigation. It was derived from a land sat 8 satellite image from the year 2019.
Distance from water sources
The DEM 30m*30m unit extension was used to construct and categorize irrigable lands near irrigation sources in a straight line (Euclidean) distance as of the watershed outlet. Following that, recatagorized distances were used for a weighted overlay investigation that included additional aspects. Calculation of possible irrigable schemes based on irrigation suitability factors. Suitability models were created by model builder within Arc tools box with tools from the spatial analyses toolsets in order to find an optimal place for surface irrigations. Following that, their appropriateness was assessed using irrigation suitability factors such as soil, land use, land cover, slope, and distance from the source as inputs to irrigation suitability models in order to identify the most suitable land for surface irrigations. The Pairwise comparison matrices approach was used to estimate the value of each parameter.
Matrices of Pairwise Comparisons Method
For balancing the standards and estimating the options for each benchmark, pair-wise comparison matrixes have been a typical technique in multiple criteria decision making. However, if the consistency ratio (CR) is less than 10%, comparison matrixes are regarded to be acceptable consistency (Saaty, 1980).
$$\text{C}\text{R}= \frac{\text{C}\text{I}}{\text{R}\text{I}}$$ ………………………………… 1.1
Where: CR is consistency ratio, CI is consistency index, RI is random index.
Calculating Water Demands for Irrigation
The total water demands of crops are the total amount of water required for the entire growing time of the area, and they are calculated using the CROPWAT8 computer algorithms, which were used to estimate mature inside the irrigation study areas. For the resolution of crop water and irrigation water needs, a model requests climatic, crop, and soil data. The model requires climate data, including average monthly smallest with greatest temperature (°C), relative humidity (percent), wind speed (km/day), and sunshine hours to calculate ETO values (hr). The FAO Penman Montieth methodologies are used to build a reference evapotranspiration ETO of single agro ecological units, utilizing the FAO CROPWAT 8 decision support software (1998). The FAO CROPWAT tool integrates procedures for determining crop evapotranspiration as well as crop irrigation demands and allows for crop water application simulations using a variety of climate data, crop varieties, planting dates, and soil conditions. Kc values for yearly expansion periods in the early, mid, and late seasons with seasonal crops were used, and these were derived from current data from FAO.
Maize, potato, barley, wheat, and cabbage were the most common crops grown in the research areas. The planting days for maize, wheat, and potato were chosen in such a way that they coincided to a local cropping calendar near the metrological station. Following that, ETO and other climate data were used to calculate crop-water requirements.
The optimal crop coefficients for these crops are determined by a number of factors (Allen et al., 1998). Because of application efficiency of 60% for surface irrigations and water conveyance efficiency of 85% for lined canals agreeing to (FAO, 1989) as the source to selected command areas, gross irrigation irrigate needs of the crops at the known possible irrigable site were computed by using the following equations (Allen and Pruitt, 1991).
ETC = ETO*Kc …………………………………………………………………… (1.2)
Where: ETc is Crop evapotranspiration (mm/day)
NIR = ETC-Pef ……………………………………………………………… (1.3)
Whereas: NIR is net irrigation requirements of crops of, Pef is effective rainfall (mm)
GIR = \(\frac{\text{N}\text{I}\text{R}}{\text{E}\text{a}}\)……………………..…………………………………………….............. (1.4)
Where; GIR is the gross irrigation requirements of crops and Ea represent the efficiency of the irrigations application within proportion. Irrigation application efficiency articulates a proportion of the amount of irrigate applied efficiency for growth of a crops. Efficiency often describes within a range of 50 percent capable of 70 percent below surface irrigation systems in Ethiopian a benchmark of the surface irrigation systems. Ea is computed at 65 percent for surface irrigations for evaluating gross irrigation requirements in this research area.
HBV Model
Because of the inputs, the HBV model is a conceptual hydrological model that simulates daily flow rate utilizing precipitation, temperatures, and possible evaporations. Prior to run-off prediction, stream flow is used to calibrate, confirm, and improve models. Fig. 1.2 Graphic constructions of the HBV models
Table 1.3 HBV model parameters
Parameter
|
Explanation
|
Unit
|
Soil Routine
FC
LP
BETA
|
highest Soil Storage (Storage in the soil)
Soil humidity values beyond which Etact reach ET
The constraint that find out a comparative condition to run-off
as of rainwater or snowmelts
|
mm
|
Response Routine
PERC
UZL
K0
K1
K2
|
highest infiltration rate as of higher to lesser region
The threshold for KO outflow
Decline coefficients (higher storage) Snow
Decline coefficients (higher storages)
Decline coefficients (lesser storages)
|
mm/ day
mm
1/day
1/day
|
Routing Routines
MAXBAS
|
Duration of the triangular weight functions
|
|
Areal rainfall of gauged Fetam watershed
The rainfall of the Fetam watershed was estimated using the Thiessen polygon method at four rainfall stations located inside and outside the watershed, namely Tilili, Shendi, Wegdade, and Kessa. After that, the areal rainfall was determined by weighting the different stations by their respective area, and the result is as follows:
P=\(\frac{1}{A}\sum _{i=1}^{n}AiPi\)………………………………………………………………………… (1.5)
Where; P is the areal precipitation, n is the amount of rain gauge station, A is the total area of the catchment Ai and Pi are the area of polygon and precipitations of each station respectively. The daily areal rainfall of the gauged Fetam watershed is calculated using this method for the years (1990 to 2004).
Figure 1.3 Thiessen polygons map of the Fetam watershed station
Potential Evapotranspiration
Using the RStudio software, the model's potential evapotranspiration input was calculated. The essential statistics for evaluating potential evapotranspiration include monthly peak and lowest temperatures, as well as rainfall (ETO).
Analyses of HBV model parameter sensitivity
Sensitivities research examines how a difference (uncertainty) in the construction of mathematical models should be allocated, qualitatively or quantitatively, to various sources of difference in the models' input (Wikipedia, 2011). It's used to calculate the change in model outputs as a function of model inputs. It is defined as the yield variable's response to changes within a contributing factor, with the bigger the change in an output response corresponding to higher sensitivity. Sensitivity analysis is useful for identifying and ranking factors that have a significant impact on certain model outputs of interest (Marian et al., 2000). Sensitivity assessments involve analyzing the performance of a model's routine in light of one or more additional restrictions that may be particularly receptive or unreceptive to changes in values. A factor that was insensitive to changes in values can be set to static values, while booming out optimizations to reduce the number of restrictions can be optimized, ensuring superior convergence to the objective task's optimum value.
HBV model calibrations and validations
The HBV Light hydrologic models' guidebook calibrations were completed by optimizing the factors that various procedures of the HBV Light hydrologic models rely on current watershed appearances. The entire 15-year period of data (1990–2004) was used in model calibrations and validations in collaboration. At this point, the first two years of statistics (1990–1991) were used to temper the start of models, followed by the remaining nine years of statistics, of which 70% data periods (1992–2000) were used for model calibrations and 30% data periods (2001–2004) were used for validation functions for Fetam watersheds.
Evaluation of the model's performance
Nash Sutcliffe Efficiency values between 0.6 and 0.8 show fair to good performance and a model is meant to be extremely good while Nash Sutcliffe Efficiency is greater than 0.8 shows extremely good performance and a model is supposed to be extremely good (Moriasi et al., 2007). PBIAS is a comparison differentiation between measured with simulated streams and actions the trend of mean simulated flows to be 10 larger or less than observed flows (Gupta et al., 1999) R2 was used to calculate the associations' goodness of fit.
Nash Sutcliffe Efficiency
Within a degree, the efficiency of a model ranges from negative infinite to best fit and is stated as;
$$\text{E}\text{N}\text{S}=1-\frac{\sum {(\text{q}\text{s}\text{i}-\text{q}\text{o}\text{i})}^{2}}{\sum {(\text{q}\text{o}\text{i}-\text{q}\text{o})}^{2}}$$…………………………………1.6
Where; qoi is the observed values on i time interval, qsi is the simulated values on i time interval, qo is the mean values of observed flow.
Coefficient of Determination- R 2
The coefficients of determination range from 0 to 1, with a 0 life form requiring no correction and a 1 life form requiring no correction. In general, values greater than 0.5 are regarded as acceptable (Legates and McCabe Jr, 1999) and should be investigated further.
\({R}^{2}\) = \(\frac{\sum _{i=1}^{n}\left(qsi-qsav\right)(qo-qoav{)}^{2}}{\sum _{i=1}^{n}{\left(qsi-qsav\right)}^{2}\sum _{i=1}^{n}{\left(qoi-qoav\right)}^{2}}\) ……………………………………………… (1.7)
Percent Bias
Positive PBIAS values imply that models are underestimating calculated values, whereas negative values indicate overestimation. Values between 10 and 15 on the discharge PBIAS scale indicate a good model simulation, whereas values greater than 25 indicate an inadequate model simulation (Linard et al., 2009) and are indicated as.
$$\text{P}\text{B}\text{I}\text{A}\text{S}=\frac{\sum _{\text{I}=1}^{\text{n}}(\text{q}\text{o}\text{b}\text{i}-\text{q}\text{s}\text{i})}{\sum _{\text{i}=1}^{\text{n}}\left(\text{q}\text{o}\text{b}\text{i}\right)}$$ ……………………………………………………………1.8
Figure 2.5 shows the overall methodology of the study.