Evaluating Land Suitability for Surface Irrigation Under Changing Climate in Gardulla Zone, Southern Ethiopia

Abstract

Climate change substantially affects water resources and agriculture, highlighting the critical importance of assessing land suitability for surface irrigation. This study was initiated with the objective of assessing the present and future land suitability for surface irrigation in the Gardulla Zone of Southern Ethiopia, utilizing meteorological, topography, soil, land cover, and proximity data. The analytic hierarchy process and weighted overlay analysis were employed to assign factor weights, while future climate projections were downscaled via a statistical downscaling model (SDSM4.2) under the shared socio-economic pathways (i.e., SSP2-4.5 and SSP5-8.5) scenarios. Irrigation suitability mapping was performed via inverse distance-weighted interpolation. The results revealed that 8% of the area is highly suitable, 54.3% is moderately suitable, 30% is marginally suitable, and 2.3% is unsuitable under current climate conditions. In the future periods, under both SSP scenarios, highly suitable land increases (up to 9.7% and 10.3% by 2050s and 10.8% and 13.5% by the 2080s under SSP2-4.5 and SSP5-8.5, respectively), whereas unsuitable land decreases (down to 0.6% by 2080s under SSP5.8.5). In terms of area, highly to moderately suitable land expanded by 1357.6–6867.7 ha, depending on the scenario and timeframe. The study concludes that climate change is expected to affect the suitability of land for surface irrigation potential in the study area and similar hydroclimatic settings, highlighting the need for forward-looking policies and adaptation options. Therefore, it is recommended to promote climate-smart irrigation systems by integrating site-specific suitability mapping into regional land-use planning and prioritizing investment in small-scale, community-managed surface irrigation schemes that reduce water losses and ensure long-term agricultural sustainability.

1. Introduction

Irrigation is the artificial application of water to agricultural land. It is a critical practice for sustaining crop productivity, particularly in the face of growing global food demand and increasing pressure on water resources [1]. As climate variability and change intensify, shifts in rainfall patterns and rising global temperatures have exacerbated water scarcity, with nearly half of the world’s population now facing severe shortages of water for agricultural use. Despite occupying only 25% of global agricultural land, irrigated agriculture contributes approximately 40% of total food production, underscoring its vital role in global food security [2]. In Africa, irrigation is not only a key driver of agricultural productivity but also a crucial component of socio-economic development and livelihoods. However, its full potential remains largely unrealized, particularly owing to the adverse influences of climate variability, comprising recurrent droughts and irregular rainfall, which constrain both existing irrigation systems and future expansion [3]. In sub-Saharan Africa, climate change is projected to further reduce the extent of agriculturally suitable land due to increased rainfall variability and extreme weather events [4]. Empirical studies in countries such as Ethiopia, Uganda, Ghana, and Morocco indicate that climate-induced alterations in rainfall and temperature are likely to alter the spatial distribution and extent of irrigable areas in the coming decades [5,6]. Ethiopia, in particular, possesses vast untapped potential for irrigation development. It is estimated that approximately 5.3 million hectares (ha) of land are suitable for irrigation, offering a strategic opportunity to increase agricultural resilience and food security under climate change [7].

Despite this potential, only a small fraction (<5%) of this land is currently irrigated, and crop production relies mainly on rainfall rather than irrigation [8,9,10]. The country often faces a decline in food production [3], which in turn cannot meet the nation’s plans to achieve the goal of agricultural development-led industrialization by 2025. Several studies have highlighted the need for alternative strategies to increase agricultural output, as the country’s traditional rain-fed farming methods cannot ensure food security [11,12]. Specifically, in the southern region, some research has been conducted on irrigation potential assessments to address the growing challenges of food insecurity [13,14,15]. Previous studies have reported that food insecurity is a significant challenge, as agricultural production in this region is primarily rainfed. Like other parts of Southern Ethiopia, the Gardulla zone is an area with land suitable for agriculture [16].

Nevertheless, there are areas in Gardulla that can be used for irrigation; however, the sector’s performance is poor, as the majority of people also use rain-fed agricultural land, which is unreliable in terms of when it starts and ends [16]. The continued reliance on unsustainable water sources, limited irrigation use for agricultural production, poor land management practices, and the use of land that is not scientifically selected on the basis of its suitability for surface irrigation exacerbate the situation. These problems have social, economic, and environmental impacts, leading to food insecurity, economic decline, and environmental degradation if left unaddressed [16].

Monitoring spatial and temporal challenges in irrigation and water resource management is more effective in addressing the issue of food security by providing information on land suitability [14]. Therefore, establishing irrigation systems that utilize available land and water resources, on the basis of the land suitability for specific purposes, is crucial, as it contributes to greater land utilization and increased agricultural output. The use of a geospatial information system (GIS)-based multi-criteria evaluation (MCE) enables a comprehensive land suitability assessment for surface irrigation [17]. It can be achieved by considering multiple biophysical and socio-economic factors that influence irrigation development [12]. The evaluation and prioritization of these suitability factors should be conducted via the AHP method [18]. Additionally, information on regional and local level climate change is crucial for determining its impacts on human and natural systems, as well as for designing suitable adaptation and mitigation strategies [19,20]. For the hydrological sector study, the statistical downscaling model (SDSM) is a suitable technique for projecting climate data in order to produce future climatic parameters [6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21].

The Gardulla Zone in Southern Ethiopia is a climatically sensitive and agriculturally important region with diverse topography and a growing population. Despite its potential for surface irrigation, it remains underrepresented in irrigation research. Increasing climate variability poses challenges to water availability and agricultural productivity, making Gardulla an ideal case for climate-resilient land suitability assessment. This study addresses a critical gap by offering localized insights that can inform sustainable irrigation planning in a similar agro-ecological region. The only available study for irrigation development in the study area, to the best of our knowledge, is Legesse and Tadesse [22], which assessed the performance of the Gatto and Arguba irrigation schemes on the basis of selected performance indicators such as water delivery performance, conveyance efficiency, and maintenance indicators. However, in their study [22], less attention has also been given to evaluating the scope and geographic extent of suitable land for irrigation via GIS-based approaches, considering a broader range of factors that hinder the development of surface irrigation in the study area. The current and future suitability of land for surface irrigation, and also the rate of change in land suitability under the impact of climate change over time, has not been adequately assessed or visualized to date.

Therefore, this study aims to present a pioneering GIS-based land suitability assessment for surface irrigation in the Gardulla Zone of Southern Ethiopia, uniquely integrating spatial analysis with climate change projections. Unlike previous assessments that rely on static environmental parameters, this research incorporates dynamic climate variables, such as shifting rainfall patterns and temperature regimes, into the suitability modeling process. By leveraging high-resolution geospatial datasets and multi-criteria evaluation techniques, the study identifies optimal irrigation zones under both current observed climate data (1991–2021) and future climate (i.e., SSP2-4.5 and SSP5-8.5) scenarios. This approach not only enhances the precision of land use planning but also provides a robust framework for climate-resilient agricultural development. The innovation of this study lies in its hybrid modeling approach that integrates climate-induced projections of water availability, land productivity, and irrigation suitability criteria, incorporating new variables that reflect future environmental shifts in the Gardulla zone, previously underrepresented in such forward-looking analyses.

2. Materials and Methodology

2.1. Description of the Study Area

The study area is located in the South Ethiopia Regional State (Figure 1) in South Omo and features diverse topography ranging from highland escarpments to semi-arid lowlands, with elevations ranging from 828 to 2617 m above sea level. The region experiences bimodal rainfall (600–1200 mm annually) and temperatures from 14.3 to 33 °C. Dominant soil types: vertisols, cambisols, and luvisols vary in water retention and agricultural suitability. Increasing climate variability has disrupted rainfall patterns, posing challenges to traditional rain-fed farming systems.

Despite its agricultural potential, Gardulla remains underserved in irrigation infrastructure and research [22]. According to the projected population CSA [23], the expected population of the town is 212,899, of which 103,770 are men and 109,129 are women. Therefore, the rapidly growing population and agricultural dependence put huge pressure on land resources, causing ecological sensitivity marked by soil erosion, deforestation, and biodiversity loss. This, in turn, makes food insecurity worse by lowering agricultural productivity and limiting access. This study applies GIS-based land suitability analysis to identify surface irrigation zones resilient to climate change, supporting informed planning and resource use in a region with significant development needs [24].

2.2. Databases

Meteorological data (i.e., rainfall, minimum and maximum temperatures) collected from the National Meteorological Institute of Ethiopia (NMI) for the 1991─2023 period for nine stations in and around the Gardulla zone. The soil property data (type, depth, drainage, and texture) at 250 m resolution were sourced from the International Soil Reference and Information Centre (ISRIC) and the Africa Soil Profiles Database (https://data.isric.org/). The land cover data were derived from Sentinel-2A imagery (from January 2024) via the Copernicus Data Space Ecosystem (https://dataspace.copernicus.eu/browser, accessed on 1 June 2024). A 30-m resolution digital elevation model (DEM) from the United States Geological Survey (USGS) was used to generate slope, elevation, and stream maps. Climate projection data collected from the Canadian Climate platform (https://climate-scenarios.canada.ca/?page=pred-cmip6#table-1/, accessed on 1 July 2024). Sample point data were collected via a field survey and Google Earth Pro for model validation. Road network data were obtained from the Gardulla Zone Road and Transportation Office for proximity analysis.

2.3. Data Pre-Processing

Data pre-processing was an important first step in ensuring accuracy in the land suitability analysis. Errors from various sources were corrected before use. The DEM data were subjected to sink filling and correction to generate slopes, river networks, and other suitability variables. Sentinel-2A imagery (January, <5% cloud cover) was used for land cover mapping because of its high spatial resolution and frequent revisit rate. Minimal processing, mainly image enhancement and color composite, was needed, as the image was already atmospherically corrected and captured under clear-sky conditions.

Daily rainfall, minimum, and maximum temperatures were collected for the 1991─2021 period, depending on the availability of data. Six meteorological stations were selected for this research on the basis of the percentage of missing data. If the recorded data for the total length of the record were approximately 80% full, the station was used in the study. Sémou et al. [25] suggested that stations with less than 20% missing values in the rainfall and daily temperature data can be used. The missing values of meteorological station data records are filled by using the inverse distance weighting (IDW) technique. It utilized observed values from other stations, as outlined in the following formula [26], to estimate missing values.

$$\mathrm{I}\mathrm{D}\mathrm{W}={\displaystyle \frac{{\sum }_{i=1 }^n\left(\frac{Vi}{Di}\right)}{{\sum }_{i=1 }^n\left(\frac{1}{Di}\right)}}$$

(1)

where vi represents the value of the same parameter at the ith nearby weather station, and Di represents the distance between the station with incomplete data and the station with a known observation.

2.4. Statistical Downscaling (SDSM)

Global climate models (GCMs) lack sufficient spatial resolution for small-scale hydrological and impact studies, making them unsuitable for regional and local assessments [27]. To address this, the statistical downscaling model (SDSM) links large-scale climate variables to local conditions, producing high-resolution projections [28]. SDSM v4.2.9 was used to downscale daily rainfall and temperature data from six stations (1991–2021) collected from the National Meteorological Institute (NMI). The climate projections were based on the higher-resolution fifth-generation Canadian Earth System Model (CanESM5: 1979–2100) experiment (CMIP6) under the “SSP2-4.5 and SSP5-8.5” emission scenarios. Twenty-six predictors from reanalysis data (1979–2014) were selected via correlation analysis for each station’s weather parameters. The model was calibrated via unconditional methods for temperature and conditional methods for precipitation at a monthly scale. Calibration and validation followed established procedures [26,27], with quality control measures applied [29] to ensure data reliability. The resulting calibrated models were used to generate synthetic weather series for current and future climates.

Both climate scenarios were selected to represent extreme and moderate climate conditions, respectively. SSP5-8.5 serves to stress-test systems under severe impacts such as drought and heat extremes [29,30] with 8.5 W/m² radiative forcing and up to 5.57 °C warming by 2100. SSP2-4.5, which projects 4.5 W/m² forcing and 1.27–3 °C warming, allows assessment of feasible mitigation and policy measures [31]. Other scenarios were excluded due to limitations: SSP1-1.9 and SSP1-2.6 assume that rapid decarbonization is unrealistic for low-adaptation regions [32]; SSP3-7.0 complicates the precipitation patterns critical for agriculture [30]; and SSP4-6.0′s focus on inequality offers less balanced policy insights than SSP2-4.5 does [33].

2.5. Data Analysis

In this study, land suitable for surface irrigation was evaluated using a GIS-based multi-criteria approach [34], as shown in Figure 2. Key parameters determining irrigation appropriateness were considered, including current and predicted climate variables (rainfall and temperature), soil characteristics (depth, type, texture), drainage, land cover, slope, and proximity to roads and rivers.

A.

Proximity and Slope

The slope of the study area was calculated to identify the suitability of an area under different slope gradients. Steep-sloped areas are not recommended for surface irrigation because they are prone to excessive runoff and soil erosion [22]. Therefore, in this study, the slope factor suitability for surface irrigation was identified by reclassifying the slope in % of Gardulla into four different suitability classes. Then, in conjunction with other factor maps, it was applied to weighted overlay analysis to identify appropriate locations for surface irrigation.

To determine the distance between a certain potential surface irrigation region and proximity factors, a proximity analysis was carried out using the Euclidean distance approach. To identify and locate suitable areas for irrigation, their proximity was determined on the basis of the study of [12,15]. Weighted overlay analysis was then performed using the reclassified distance in conjunction with other suitability analysis factors.

B.

Land Cover (LC)

The study area’s land cover was classified on the basis of irrigation suitability via supervised classification of Sentinel-2A imagery. This method, supported by true-color composites and false-color combinations, was chosen for its higher accuracy over unsupervised approaches [35]. The land cover types identified included built-up areas, water, forests, bare land, grazing land, and cultivated land. These classes were categorized into four levels of suitability: unsuitable (N), slightly suitable (S3), suitable (S2), and highly suitable (S1) following Kassaye et al. [32]. Eight hundred forty-six training points and 450 reference points (≥30 per class), sourced from Google Earth, ensured reliable classification accuracy. The supervised classification achieved 89.56% overall accuracy and a kappa coefficient of 0.86. However, the producer accuracy for built-up and bare land [36] was lower due to class imbalance and spectral similarity with other land cover types.

C.

Soil and Climate Factors

The suitability of soil for irrigation depends on properties such as texture, fertility, moisture content, depth, and drainage, which influence productivity and irrigation performance [37]. This study assessed the depth, texture, type, and drainage of the soil as they affect the infiltration rate, nutrient retention, water movement, susceptibility to waterlogging and runoff, organic matter content, root penetration, and water-holding capacity [6,22]. ISRIC-AfSIS soil data, available at depths of 0–200 cm with a 250 m resolution, were used for analysis [17]. The vector-format soil data were rasterized and reclassified via the spatial analysis reclassify tool to assign suitability scores. Rainfall and temperature are included to reflect an area’s vulnerability to moisture stress under a changing climate [14,17,22]. Climate data was downscaled using SDSM to generate localized projections of climate variables such as temperature and precipitation that were then used as input criteria in the AHP framework with other factors. The daily climate data were combined into baseline, 2050, and 2080 mean annual temperature and precipitation data. Raster layers are generated through inverse distance-weighted (IDW) interpolation, an accurate geospatial technique for estimating unsampled values [37]. Interpolation settings were adjusted to match the study area extent, using both in-area and nearby weather stations. All raster layers were reclassified and integrated via weighted overlay analysis to identify irrigation-suitable areas.

2.5.1. Standardization of Suitability Criteria

In order to integrate all datasets, firstly, all vector datasets were converted to raster format. The nearest neighbor method was used to resample all datasets to a resolution of 30 m × 30 m. Then, in accordance with the land suitability classification rules [37], each element was classed into four land suitability classes: highly suitable (S1), moderately suitable (S2), marginally suitable (S3), and not suitable (N). Suitability maps were generated for each factor. Following this, weights have been assigned for each factor using the MCDA-AHP technique, and further weighted overlay analysis was performed by integrating all the raster dataset layers in the GIS environment to generate a suitability map, which was then validated using ground-truth data for accuracy assessment. Under all climate scenarios, highly suitable (S1) areas require significant irrigation due to high water demand and increased susceptibility to moisture deficits under climate change [22,34]. Moderately suitable (S2) areas are moderately affected and need additional irrigation. “Marginally suitable (S3)” areas generally receive sufficient rainfall and may not require irrigation, whereas “not suitable (N)” areas typically have adequate natural moisture and do not need irrigation [14,22].

2.5.2. Multi-Criteria Factor Weighting Through the AHP

As part of the multi-criteria decision analysis (MCDA) framework proposed by Saaty [38], this study used a GIS-based analytical hierarchy process (AHP) to identify the optimal land for surface irrigation. The AHP was selected for its ability to prioritize criteria in complex decision making systematically and to translate expert judgments into numerical weights for spatial analysis [18]. The relative relevance of the factors was evaluated using a pairwise comparison matrix based on Saaty’s 1–9 importance scale [38]. Weights were obtained from expert input, a literature review, and the local environmental context. A purposive sampling approach was used in which experts were chosen based on specific criteria related to their work experience, academic knowledge, and expertise. In total, 16 experts participated in the AHP analysis. These included specialists in water resources and irrigation systems from the Water and Energy Office of the study area, a soil and agronomy expert from the Agricultural Office of Gardulla with expertise in crop production and soil management, and experts in land management and policy, climatology, and GIS from the Agricultural Office. This diversity of backgrounds ensured representation of key perspectives relevant to land suitability for irrigation. Furthermore, we cross-validated the expert judgments with findings from the literature and the local environmental context, thereby enhancing the reliability of the weights and minimizing subjectivity. To identify suitable irrigation areas, the weighted parameters were combined using weighted overlay analysis [31]. To validate the pairwise comparisons, the consistency ratio (CR) was computed, with a CR < 0.10 indicating acceptable consistency, following Saaty’s method [39].

SDSM-downscaled rainfall, minimum, and maximum temperature data can be integrated into land suitability analysis for surface irrigation by using the Analytic Hierarchy Process (AHP). AHP is used to construct pairwise comparison matrices, derive weights for each criterion, and ensure consistency in judgments. Finally, a weighted linear combination of all standardized criteria produces the overall surface irrigation suitability map, which can be classified (e.g., highly suitable, moderately suitable, marginal, or unsuitable), with the added advantage of analyzing both current and future climate scenarios provided by SDSM.

2.5.3. Surface Irrigation Suitability Under Present and Future Climates

Surface irrigation suitability was assessed for both the current (1991–2021) and future periods under climate change. The current suitability map was developed using observed temperature, rainfall, and other biophysical factors. Future suitability was evaluated under “SSP2-4.5 and SSP5-8.5” climate scenarios for the middle century (2050: 2035–2065) and late century (2080: 2066–2096), with climate projections downscaled via the SDSM. The downscaled rainfall and minimum/maximum temperatures were aggregated into annual values, rasterized, and reclassified for suitability analysis. Future suitability maps were generated by substituting current climate layers with projected data, whereas soil, slope, land cover, and river layers were assumed to be static [6]. The shifts in suitability due to climate change were identified by comparing current and future maps. Prior to weighting, a “constraint map” was developed to disregard no irrigable parts such as protected forests, lakes, and urban regions, following the approach of Wanyama et al. [22]. The AHP-based weighting analysis was repeated for the 2050 and 2080 scenarios to evaluate changes in irrigation suitability under projected climatic scenarios.

The baseline map and projected suitability maps for 2050 and 2080 (under SSP2-4.5 and SSP5-8.5) were compared in order to evaluate changes in land suitability for surface irrigation brought on by climate change. The suitability shifts were quantified by subtracting the current from the future suitability values, and the percentage change was calculated relative to the baseline. Positive values indicate increased suitability and greater irrigation demand, whereas negative values reflect declining suitability due to unfavourable climatic conditions. Identifying these shifts will support adaptive land and water resource management under changing climate scenarios.

2.5.4. Validation of the Land Suitability Analysis

The irrigation suitability map under climate change was validated in this study using the ROC curve. The ROC method, which is grounded in established statistical principles, offers a robust and standardized approach for accurate performance evaluation [40].

3. Results

3.1. Assessing the Current Suitability of Land for Surface Irrigation

This study analysed surface irrigation suitability by considering different land suitability evaluation factors, such as roads, rivers, drainage, LC, precipitation, temperature, slope, soil type, texture, and depth.

3.1.1. Slope Factor

Slope is considered a key parameter in evaluating surface irrigation suitability, as it significantly influences irrigation method efficiency, erosion risk, and land usability. The study area’s slope was categorized into four suitability classes [38,41]: 0–2% (S1: extremely suitable), 2–5% (S2: moderately suitable), 5–8% (S3: marginally suitable), and >8% (N: not suitable). Suitability was assessed according to slope steepness. The results revealed that, of the 69,938-ha total area, 43.4% (30,377.6 ha) were highly suitable, 31.3% (21,916.4 ha) were moderately suitable, 18.4% (12,901 ha) were marginally suitable, and 6.8% (4742.9 ha) were not suitable for surface irrigation. According to Figure 3b, roughly 74.7% of the land is in the highly to moderately suitable category, indicating that gentle slopes that are conducive to surface irrigation development and agricultural usage predominate.

Steep slopes restrict the remaining 25.2% of the land’s ability to be irrigated. The western, northeastern, and southern regions are primarily occupied by areas with slopes less than 2%, where infiltration is efficient, allowing irrigation without major restrictions. Slopes of 2–5%, found centrally and from the northwest to southwest and parts of the east, may cause some runoff due to gradient variation, yet still permit reasonable infiltration. Land with 5–8% slopes, located in central, eastern, and southwestern zones, poses greater runoff risk and reduced infiltration but may be adapted for irrigation with interventions such as land levelling or terracing. Areas exceeding 8% slope, mainly in the central, southwestern, and eastern parts, are unsuitable for surface irrigation, which aligns with FAO recommendations [42].

Overall, the Gardulla Zone is largely characterized by gentle slopes, making it more viable for surface irrigation, which is especially critical given the region’s high dependence on rainfall. These findings align closely with earlier studies [43], which estimated that approximately 74.4% of the area supports cultivation due to its gentle topography, low erosion risk, and high suitability.

3.1.2. Soil Factors

Soil characteristics such as texture, drainage, type, and depth were considered in this analysis. Among these, soil texture is essential for evaluating infiltration and water-holding capacity, both of which are critical for irrigation suitability. On the basis of previous studies [12,38,40], four texture classes (d) were identified in the study area: clay (15.1%), clay loam (73.5%), sandy clay loam (7.6%), and sandy clay (2.7%). Clay soils were rated as highly suitable (S1) owing to their moisture and nutrient retention, whereas clay loam soils were moderately suitable (S2). Sandy clay loam soils, with lower organic matter and water retention, were marginally suitable (S3), and sandy clay soils were deemed unsuitable (N) because of their poor water-holding capacity. A 0.83% portion of the area occupied by Lake Chamo was excluded because of missing data.

Overall, 10.3% of the land is categorized as marginally suitable or not suitable for surface irrigation, while 88.6% of the land is characterized as highly to moderately suitable. This finding highlights the favourable distribution of texture classes in the region, particularly the dominance of clay loam, which supports sustainable irrigation development with moderate inputs. The results presented here concur with Getahun et al. [12] findings, who reported similar suitability levels based on texture in Ethiopian agro-ecological settings. These findings emphasize the significance of considering soil texture in surface irrigation and support targeted interventions, such as water conservation techniques in marginal areas, to optimize land productivity under limited water availability.

Soil drainage: On the basis of the FAO soil drainage classification [40] following the study of [14], the study area revealed four primary drainage categories (Figure 4a): well-drained, moderately drained, imperfectly drained, and poorly drained soils. Soils that drain well, which account for 81.5% (57,010.4 ha) of the region, were categorized as highly appropriate for surface irrigation due to their ideal water retention and infiltration properties. Moderately drained soils, accounting for 9.5% (6631.1 ha), were considered moderately suitable and were mostly found in the southern farmlands of Elbache and northeastern Holte. Imperfectly drained soils (4.8%, 3355.2 ha) were deemed marginally suitable because they poorly manage excess surface water, whereas poorly drained soils (4.2%, 2941.3 ha) were classified as not suitable because of restricted water movement. Overall, 91% of the area is classified as highly to moderately suitable for surface irrigation via drainage, indicating favourable conditions for agricultural development, which is consistent with findings of Girma et al. [14], who reported that 98.7% of their study area was similarly suitable.

Soil type: Soil type plays a critical role in surface irrigation suitability because it influences organic matter content, texture, water-holding capacity, depth, and salinity. In the Gardulla zone, eight major soil categories were identified: Cambisols, Nitisols, Vertisols, Leptosols, Andosols, Acrisols, Ferralsols, and Luvisols (Figure 4c). These were reclassified into four suitability classes (not suitable, marginally suitable, moderately suitable, and highly suitable) following FAO guidelines [40]. Nitisols (9.7%) and Luvisols (7.9%) were deemed highly suitable (S1) because of their fine texture, good drainage, and sufficient depth, which support optimal root development. Cambisols (23%) and Vertisols (19.2%) were classified as moderately suitable (S2) because of their moderate to deep profiles, although they present some limitations in terms of water retention or workability.

In contrast, Andosols (2.1%) and Ferralsols (2.9%) were found to be marginally suitable (S3), mainly because of limitations in water retention and fertility. In comparison, Leptosols (32.8%) and Acrisols (2.4%) were not suitable (N) because of their shallow depth and poor drainage. Overall, just 17.6% of the area was assessed as highly favorable, with 42.2% classified as moderately suitable. The remaining 39.9% fell under the marginal or not suitable categories. These findings highlight the variability of soil resources in Gardulla and emphasize the need for site-specific irrigation planning. These results are in line with findings of Girma et al. [14], which showed Vertisols and Cambisols as dominant moderately suitable soil types for irrigation in similar agro-ecological settings, highlighting the broader applicability of these findings in land and water resource management.

Soil depth: Soil depth plays a vital role in plant growth, besides root development, directly influencing land suitability for surface irrigation. The soil depth in the study area ranges from 0 to 200 cm, with Gardulla exhibiting depths between 48 and 100 cm. Suitability was determined (Figure 4a) using known thresholds: <50 cm (unsuitable), 50–80 cm (marginally suitable), 80–100 cm (moderately suitable), and >100 cm (very suitable) [6,17].

The analysis revealed that areas in western Gardulla (around Busa), northern parts (near Holte), and farmlands near Ateya and Gato are highly suitable for surface irrigation, with depths exceeding 100 cm and covering approximately 16,235.6 ha (23.2%). Moderately suitable soils occupy 39,981.2 ha (57.2%), disseminated across mainly the northwestern, central, and eastern regions. Marginally suitable soils (12,100.3 ha; 17.3%) occur adjacent to these zones. Only 2.3% (1620.9 ha) was categorized as unsuitable because of limited depth and poor water retention. Overall, the region is predominantly characterized by moderate to highly suitable conditions for surface irrigation on a soil depth basis. The four soil factor maps (texture, drainage, type, and depth) were reclassified to a common scale and integrated through a GIS-based weighted overlay to produce a composite soil suitability map (Figure S1).

3.1.3. Land Cover

Land cover is an important aspect in determining land suitability for surface irrigation since it reflects soil productivity and the need for terrain alteration. In the study area, cultivated land dominated, covering 37,726.07 ha (54.3%), followed by grazing land (19,416.3 ha; 27.8%), vegetation (8202.04 ha; 11.72%), bare land (2533.9 ha; 3.62%), built-up areas (1068.2 ha; 1.52%), and water bodies (981.7 ha; 1.04%). Girma et al. [15] classified cultivated land as highly suited for surface irrigation, whereas grazing land is moderately suitable, and barren land is marginally suitable. Furthermore, water, forests, and urban locations are inappropriate (Figure 3d).

Accordingly, cultivated land classified as highly suitable covers the majority of the Gardulla zone, indicating a high potential for surface irrigation with minimal land modification (Figure 3d). Moderately suitable land (mainly grazing areas) accounts for 19,418.76 ha (27.8%) and is distributed throughout the zone, particularly around the Gardulla forest and Gato. Marginally suitable areas (2536.3 ha; 3.6%) are mostly bare lands located in parts of Moseye, Holte, Mashole, Ateya, and Gato. The remaining 10,254.4 ha (14.7%) comprises unsuitable land cover (e.g., water bodies, dense vegetation, and paved areas). These findings show the predominance of cultivated land suitable for irrigation.

3.1.4. River and Road Proximity

River proximity is a critical factor in assessing surface irrigation potential, as access to nearby water sources reduces dependence on rainfall and enhances agricultural productivity [44,45]. In this study, suitability based on distance from rivers was classified following Tolera et al. [36] into highly suitable (0–0.735 km), moderately suitable (0.735–1.68 km), marginally suitable (1.68–2.94 km), and not suitable (2.94–6.698 km). The findings showed that 42.8% of the research area was considered highly suitable, 33.7% was moderately suitable, 18.8% was marginally suitable, and 4.7% of the land was unsuitable for surface irrigation (Figure 3c). This finding indicates that the majority (76.5%) of the Gardulla species have good access to river water and are well-positioned for surface irrigation with minimal infrastructure investment. The remaining areas, which are located farther from rivers, are less suitable because of the higher costs associated with water conveyance. Muluneh et al. [13] also reported that river proximity was a key determinant of land for surface irrigation suitability.

Road accessibility is a key factor influencing irrigation efficiency and market access, as emphasized by Tolera et al. [36] and Worqlul et al. [6]. In this study, the Euclidean distance was used to classify proximity to main roads in Gardulla, following Mandal et al. [46], with buffer zones defined as 0–4 km (highly suitable), 4–9 km (moderately suitable), 9–16 km (marginally suitable), and >16 km (not suitable). The analysis (Figure 3c) shows that 50.4% (35,214.6 ha) of the land is extremely appropriate, with the majority located in Gardulla’s central, northern, and southeastern regions. Moderately and marginally suitable areas cover 27.8% (19,474.4 ha) and 17.3% (12,070.4 ha) of the total area, respectively. In contrast, only 4.5% (3178.6 ha) of the total area is not suitable because of the greater distance from roads. These findings align with those of Dinku and Kebede [47], who highlighted that proximity to roads supports the efficient transport of inputs and outputs. Overall, over half of the study area benefits from favourable road access for surface irrigation development.

3.1.5. Climate Variable Suitability for Irrigation

When assessing the land suitability for irrigation, climate factors like temperature and rainfall are crucial. Given the semi-arid nature of much of Gardulla, regions with higher temperatures and less precipitation are thought to be more suitable for surface irrigation due to their higher moisture deficit [15]. A precipitation-based analysis showed that, while 43.4% of the land is marginally suitable and 22.2% is not suitable, 13% (9040 ha) and 21.4% (15,014 ha) of the region are extremely favorable and moderately suitable, respectively. 38% of the land is marginally favorable, 41% is not suitable, 10% is fairly suitable, and 11% is highly suitable in terms of temperature. These results suggest that, under current climate conditions, a large portion of the study area has sufficient natural moisture and thus lower irrigation demand.

3.1.6. Weight Determination for Factors via the AHP

To evaluate the suitability of land for surface irrigation, ten key factors were compared pairwise. This technique involves systematically evaluating the relative importance of each factor against others in a one-to-one matrix on the basis of the scale developed by Saaty [37]. All the factors were arranged in both rows and columns to construct the comparison matrix, from which the weights were derived (

Table 1. Weights of each considered factor.

Factors	Weights	Weight (%)
PCP (Rainfall)	0.2176	22
Temperature	0.1735	17
Soil Depth	0.1610	16
Slope	0.1054	11
Land cover	0.1008	10
River	0.0761	8
Soil Type	0.0642	6
Road	0.0462	5
Soil Texture	0.0324	3
Soil Drainage	0.0229	2
Total	1	100

The weights in were determined using the Analytic Hierarchy Process (AHP), where ten key factors were compared pairwise following Saaty’s scale. The pairwise comparison matrix was carefully constructed to reflect the relative importance of each factor for surface irrigation in the study area, and the consistency of the judgments was confirmed (CI = 0.08; CR = 0.05, which is <0.1, indicating acceptable consistency). While we recognize that the resulting weights (e.g., PCP 22%, temperature 17%) may not be directly generalizable to other regions, we minimized subjectivity by grounding the comparisons in local environmental conditions and by aligning our weighting decisions with findings from previous studies that emphasize the dominant role of climatic variables in land suitability for irrigation. Therefore, the weights are context-specific to the study area, but the method itself remains adaptable and can be recalibrated for different regions depending on local conditions.

). The results indicated that rainfall had the greatest influence on land suitability, followed by temperature, soil depth, slope, land cover, river proximity, soil type, road proximity, soil texture, and drainage. The prominence of climatic variables aligns with previous studies highlighting the dominant role of precipitation and temperature in determining land suitability for agriculture. Therefore, assigning greater weights to climate-related factors is critical when evaluating suitability under changing climate conditions. Using a random index (RI = 1.49), the consistency index (CI = 0.08) and consistency ratio (CR = 0.05) for ten factors were computed in order to confirm the consistency of the pairwise comparisons. There is an acceptable degree of consistency in the assessments when the CR value is less than 0.1. The final weights assigned to each factor are presented in Table 1.

3.2. Constraints and Current Land Suitability for Surface Irrigation

The term “constraints” describes elements that restrict the use of water and land resources for irrigation [17]. In this study, the constraint variables included protected forest areas, lakes, and urban settlements (Figure 5A). These areas were identified as permanently unsuitable for surface irrigation. Following the weighting of all relevant factors, a constraint map was applied to exclude these areas from the analysis. A surface irrigation suitability map was subsequently generated, which displayed varying degrees of suitability across the remaining landscape (Figure 5B).

About 5662.7 ha, mostly in the southern and southeast peripheries, was estimated to be very suitable for surface irrigation in Gardulla, according to the land suitability analysis (Figure 5). The majority of the region, with the exception of a few central, eastern, and southeast regions, has 37,993.2 ha (54.3%) of moderately suitable land. Marginally suitable land covered 20,998.7 ha (30.0%), whereas 3653.6 ha (5.2%) was deemed currently unsuitable due to various limiting factors (Figure 1, Figure 3 and Figure 4). It was determined that 43,655.8 ha, or 62.4% of the total land area, was either moderately or highly suitable for surface irrigation.

3.3. Validation Results of the Suitability Model

Model validation was conducted via 122 reference points derived from field surveys and Google Earth Pro to evaluate the accuracy of the suitability map. Prior to analysis, both the sample points and the raster map were projected to a common coordinate system. With the area under the curve (AUC) serving as the performance metric, validation was carried out using the receiver operating characteristic (ROC) curve. The five categorized [40] classes, 0.9–1 (Excellent), 0.8–0.9 (Very good), 0.7–0.8 (Good), 0.6–0.7 (Moderate), and 0.5–0.6 (Poor), were used to calculate the AUC values.

The AUC value was 0.84, indicating very good model performance in identifying suitable areas for surface irrigation (Figure 6). An AUC > 0.8 is generally considered indicative of a reliable model [37]. This result agrees with the findings of Ismaili et al. [45], endorsing that the predicted suitability map realistically represents ground conditions. A thorough assessment of model accuracy is provided by the ROC curve, which compares the true positive rate versus the false positive rate. On the basis of standard classification thresholds where AUC values of 0.8–0.9 are deemed very good, the model demonstrates strong predictive capability for surface irrigation suitability mapping.

3.4. Statistical Downscaling and Prediction of Future Climate Variables

The most pertinent indicators for downscaling local stations’ maximum and minimum temperatures, as well as rainfall, were found using partial correlation analysis. Among the 26 predictors, 14 were selected as significant, with temperature-related variables contributing the most (47.6%), followed by p500 (38.1%). For there to be a strong correlation between local climate conditions and large-scale predictions, model validation and calibration are crucial. The period 1991–2002 was used for calibration, and the period 2003–2014 was used for validation. The performance of the SDSM was evaluated using the coefficient of determination (R²) and the root mean square error (RMSE). A strong agreement between observed and simulated values is indicated by R² values close to 1, while lower RMSE values signify better model performance.

During the calibration phase, R² values for maximum temperature ranged from 0.9033 to 0.9945, while those for precipitation ranged from 0.8933 to 0.9879. In the validation phase, R² values for temperature and precipitation varied between 0.8907–0.9212 and 0.8986–0.9326, respectively. These results demonstrate a strong correlation between observed and simulated data, confirming the reliability of SDSM for downscaling in the study area. The RMSE values further supported this conclusion. For precipitation, RMSE ranged from 5.731 to 17.35 during calibration and from 5.212 to 8.765 during validation. For temperature, RMSE values ranged from 0.47 to 1.031 during calibration and from 0.562 to 0.924 during validation. Together, these performance metrics confirm that the SDSM adequately captures the variability of both temperature and precipitation, consistent with findings from previous studies that applied SDSM in diverse climatic regions (e.g., [19,21,47]).

3.5. Future Climate Suitability Classification for Surface Irrigation

3.5.1. SSP4.5 Scenarios

Table 2. Reclassified future climate suitability for surface irrigation under the SSP8.5 scenario.

Scenario	Suitability Class	Pcp2050		Temp2050		Pcp2080		Temp2080
		ha	%	ha	%	ha	%	ha	%
SSP4.5	S1	9536	14	8221	12	10,110	14	9384	13
	S2	13,532	19	9243	13	22,250	32	12,113	17
	S3	31,856	46	30,032	43	27,679	40	37,286	53
	N	15,014	21	22,442	32	9899	14	11,155	16
	Total	69,938	100	69,938	100	69,938	100	69,938	100
SSP8.5	S1	10,044.9	14.4	9013	12.9	10,509.6	15	9474.9	13.6
	S2	16,168.3	23.1	11,871.7	16.9	22,595.3	32.3	14,501.7	20.7
	S3	33,748.5	48.3	35,646.8	50.9	26,858.3	38.4	35,425.9	50.7
	N	9976.3	14.3	13,406.4	19.2	9974.73	14.3	10,535.4	15.1
	Total	69,938	100	69,938	100	69,938	100	69,938	100

presents the reclassified current and downscaled climate variables under the SSP4.5 scenario. The suitability analysis of the mean annual precipitation and temperature revealed spatially variable changes across the study area from the baseline to future periods (2050s and 2080s). In the 2050s, precipitation suitability shifted toward increased suitable areas, especially in the northwest, with the largest gains observed in the moderately and highly suitable classes. This trend intensified in the 2080s, when unsuitable areas further decreased while suitable classes expanded.

Temperature suitability also showed positive shifts in the 2050s and 2080s; areas categorized as highly and moderately suitable increased, particularly in the southern and southeastern parts. However, marginally suitable areas expanded in some zones during the 2050s, especially in the northwest, but declined by the 2080s due to continued decreases in warming and rainfall. The spatial maps (Figure 7) confirm these patterns, showing reduced unsuitable land and increased suitability, especially in the central, eastern, and western areas.

Overall, the results indicate significant shifts in surface irrigation suitability under climate change. Although suitability has improved in many areas, the rate and extent of change vary between the 2050s and 2080s, with the latter experiencing more pronounced shifts. The areas that are predicted to be most impacted are the eastern, southern, and partly western zones. These findings align with projections by Alemayehu et al. [48], who anticipate rising temperatures across Ethiopia under similar climate scenarios. This emphasizes how crucial it is to take account of climate projections in land use and irrigation planning.

Figure 7 shows the reclassified distribution of the observed mean annual precipitation (Figure 7A) and temperature (Figure 7D), PCP (Figure 7B) and temperature (Figure 7E) of the 2050s, and PCP (Figure 7C) and temperature (Figure 7F) of the 2080s under the SSP4.5 scenario. The spatial distribution of observed annual average temperature and mean annual PCP suitability is similar to that of 2050 suitability, but there is a further decrease in unsuitable areas and an increase in highly, moderately, and some marginally suitable areas. Most of the land that is highly suitable for surface irrigation in the observed period is still found in the southern parts of the study area in both future periods. Moderately suitable lands are found in all parts except some central, southern, and western areas.

3.5.2. SSP85 Climate Suitability

Under the SSP8.5 scenario, the projected rainfall and temperature suitability analysis revealed increases in highly suitable land areas to 10,044.90 ha and 9013.04 ha, respectively, with expansions of 1004.67 ha and 1031.95 ha, respectively, matched with the reference period in the 2050s. This intensification is expected to continue, reaching an additional 1469.39 ha and 1493.90 ha by the 2080s. In terms of rainfall suitability, the largest increase in the 2050s occurred in marginally suitable land (33,748.5 ha), followed by moderately suitable land (16,168.29 ha), relative to the baseline period.

By the 2080s, the area of moderately suitable land for rainfall increased to 22,595.31 ha, whereas the area of marginally suitable land decreased to 26,858.34 ha. Compared to the baseline in the 2050s and 2080s, the extent of land deemed unsuitable for surface irrigation decreased to 9976.32 ha, 5552.85 ha, and 9974.7 ha, respectively. These changes reflect shifts in rainfall patterns over time. Similarly, for temperature suitability, marginally suitable land expanded to 35,646.81 ha, and moderately suitable land expanded to 11,871.74 ha in the 2050s. Unsuitable land decreased to 13,406.4 ha in the 2050s and to 10,535.4 ha in the 2080s, corresponding to reductions of 15,307.69 ha and 18,178.7 ha, respectively.

Spatially, highly suitable areas for surface irrigation under SSP8.5 remain concentrated in the southern region across both future periods (Figure 8A–C). The western, northwestern, southwestern, and some southeastern regions of Gardulla are home to slightly suitable land. In contrast, marginally suitable land expanded across the central and eastern regions, which was consistent with the SSP4.5 patterns. In the 2080s, eastern areas transitioned to moderately suitable areas under rainfall conditions, although they remained marginally suitable for high temperatures. Central areas consistently retain some suitability for both climate variables, except for temperature in the 2050s, which shifts eastwards (Figure 8C). Overall, SSP8.5 projections suggest greater climate-driven changes in land suitability, primarily due to declining rainfall and rising temperatures, than the moderate-emission SSP4.5 scenario does, which aligns with previous findings [49].

3.6. Land Suitability for Surface Irrigation Under Emission Scenarios in the 2050s

Under the SSP4.5 and SSP8.5 scenarios, displays the regional distribution of land suitable for surface irrigation in the 2050s. Marginally suitable areas and non-suitable areas are located primarily in the central, eastern, and southwestern parts of Gardulla under both scenarios. Conversely, highly suitable lands are concentrated in the southern and southeastern regions, particularly within the Rift Valley and along some western peripheries of Gardulla. This shift suggests an increased frequency of suitable conditions for irrigation compared with the baseline, which is likely driven by climate change. Moderately suitable areas are distributed across most parts of the region, except the central and eastern zones, where they increase from the baseline. Overall, 64.4% (45,013 ha) and 66.5% (46,481 ha) of Gardulla fell within the moderately to highly suitable classes under SSP4.5 and SSP8.5, respectively. This suggests a growing demand for surface irrigation due to rising temperatures. These findings align with those of Wanyama et al. [22], who projected increased irrigation requirements in future periods that may exceed available water supplies.

3.7. Land Suitability for Surface Irrigation Under Emission Scenarios in the 2080s

The projected land suitability analysis for the 2080s (Figure 9) indicates that, under the SSP4.5 scenario, moderately suitable to highly suitable land for surface irrigation would increase to 40,330.8 ha (57.7%) and 7573.6 ha (10.8%), respectively, from the baseline (

Table 3. Potential land suitability for surface irrigation in the 2080s for each scenario.

Suitability Class	SSP4.5 Scenario		SSP8.5 Scenario
	(ha)	(%)	(ha)	(%)
S1	7573.6	10.8	9432.4	13.5
S2	40,330.8	57.7	41,091.2	58.8
S3	17,850.6	25.5	15,320	21.9
N	529.4	0.8	440.8	0.6
Constraint area	3653.6	5.2	3653.6	5.2
Total	69,938	100	69,938	100

). In contrast, marginally suitable and unsuitable land would decrease to 17,850.6 ha (25.5%) and 529.4 ha (0.8%), respectively. Under the SSP8.5 scenario, the increase is even more pronounced, with highly suitable and moderately suitable land projected to reach 9432.4 ha (13.5%) and 41,091.2 ha (58.8%), respectively. Moreover, the area of unsuitable and marginally suitable land would decrease to 440.8 ha (0.6%) and 15,320 ha (21.9%), respectively. These changes in land suitability under different climate conditions are projected to significantly impact water resources by reducing surface runoff, groundwater recharge, and total water yield, which will lead to water scarcity for agriculture and other uses in societies that mainly depend on rainfall.

However, the extent of change differs between the scenarios, with the most substantial shifts observed under the SSP8.5 scenario, reflecting its status as a high-emissions pathway. Despite these changes in suitability classes, the spatial distribution of suitability remains generally consistent across the zone compared with that in the 2050s. These results are consistent with those of previous studies by Nigatu et al. [3] and Dawit et al. [48], which concluded that future climatic changes, especially by the 2080s, will significantly influence land suitability for agriculture.

3.8. Land Suitability from the Baseline Era for Surface Irrigation in the Future

The effects of climate change cause variations in the frequency of changes in the suitability of land for surface irrigation over time.

Table 4. Land suitability class for surface irrigation in the context of SSP4.5 and SSP8.5.

Suitability Class	SSP4.5 (2050)		SSP8.5 (2050)		SSP4.5 (2080)		SSP85 (2080)
	(ha)	%	(ha)	%	(ha)	%	(ha)	%
S1	1093	19.3	1538.7	27.2	1910.9	33.7	3769.7	66.6
S2	264.6	0.7	1286.4	3.4	2337.6	6.2	3098.0	8.2
S3	−583	−2.8	−1882	−9	−3148	−15	−5678.6	−27
N	−774.5	−47.5	−943	−57.9	−1100.4	−67.5	−1189.0	−73

summarizes the shifts in land suitability under the SSP4.5 and SSP8.5 scenarios for the 2050s and 2080s. In both scenarios, the area classified as highly and moderately suitable for surface irrigation increases over time, whereas marginally suitable and unsuitable areas decrease relative to the baseline period. Notably, the changes are more pronounced under the high-emission SSP8.5 scenario than under SSP4.5. In the 2050s, highly suitable land is projected to increase by 1093 ha (19.3%) under SSP4.5 and by 1538.7 ha (27.2%) under SSP8.5. By the 2080s, this increase reached 1910.9 ha (33.7%) and 3769.7 ha (66.6%) under SSP4.5 and SSP8.5, respectively. Similarly, moderately suitable land expanded by 264.6 ha (0.7%) in 2050 and by 2337.6 ha (6.2%) in 2080 under SSP4.5. Under SSP8.5, the increase is 1286.4 ha (3.4%) in 2050 and 3098 ha (8.2%) in 2080.

Conversely, marginally suitable land is projected to decline by 583 ha (−2.8%) in 2050 and by 3148 ha (−15%) in 2080 under SSP4.5. For unsuitable land, a substantial reduction is expected: by 774.5 ha (−47.5%) in 2050 and 1100.4 ha (−67.5%) in 2080 under SSP4.5 and by 943 ha (−57.9%) in 2050 and 1189 ha (−73%) in 2080 under SSP8.5. These findings indicate that the frequency and magnitude of land suitability shifts are more significant in the 2080s, particularly under the SSP8.5 scenario.

Figure 9 further illustrates the projected changes in land suitability from the baseline. While the spatial extent of change varies by period and scenario, the most substantial shifts occurred in the 2080s, driven by increasing temperatures and declining rainfall under high-emission conditions. The study’s overall findings support the prediction that climate change will significantly alter the study area’s soil suitability for surface irrigation, particularly under SSP8.5. These findings align with those of Wanyama et al. [22]. Significant changes in the suitability of land for irrigation under future climatic scenarios were also reported by Worqlul et al. [6].

In the results, we found that, under both SSP2-4.5 and SSP5-8.5 scenarios, the proportion of highly suitable land for irrigation is projected to increase compared to the baseline. Specifically, under SSP2-4.5, highly suitable land increases from the current 8% to 9.7% by the 2050s and 10.8% by the 2080s, while under SSP5-8.5 it rises more substantially, reaching 10.3% by the 2050s and 13.5% by the 2080s. These shifts indicate that climate change may expand the potential area for irrigation, especially under higher emission pathways. However, the results should be interpreted cautiously, as the expansion of suitable land does not automatically translate into sustainable irrigation development. The increased demand on water resources, particularly under SSP5-8.5, where evapotranspiration and water scarcity risks are higher, could offset the apparent benefits. Additionally, socio-economic consequences, such as land-use competition, investment costs, and equity in water access, were not explicitly analyzed in this study but remain important for policy implications.

4. Discussion

This study provides a comprehensive assessment of land suitability for surface irrigation in the Gardulla Zone under current and projected climate scenarios. The results reveal that only 8% of the land is currently highly suitable for surface irrigation, while 54.3% is moderately suitable, 30% marginally suitable, and 2.3% unsuitable. Under future climate scenarios (SSP2-4.5 and SSP5-8.5), there is a consistent increase in highly suitable land, reaching up to 13.5% by the 2080s under SSP5-8.5, while unsuitable land declines to 0.6%. These shifts are primarily driven by changes in precipitation and temperature, which were identified as the most influential factors through the AHP. This finding is consistent with Worqlul et al. [6], who emphasized the sensitivity of irrigation suitability to climate variability in Ghana. The integration of soil texture, slope, land cover, and proximity to infrastructure through a GIS-based multi-criteria evaluation provided a robust spatial framework for assessing irrigation potential, aligning with methodologies employed by Muluneh et al. [13].

The spatial analysis revealed that the southern and southeastern regions of Gardulla consistently emerged as hotspots for surface irrigation due to favorable soil conditions (clay loam), gentle slopes (<5%), and proximity to rivers and roads. These findings are supported by Girma et al. [15], who identified Vertisols and Cambisols as moderately suitable for irrigation in Ethiopia’s Omo-Gibe Basin, particularly when drainage conditions were favorable. Similarly, Muluneh et al. [13] underscored the importance of slope and river proximity in determining irrigation feasibility in the Rift Valley Lakes Basin. Alemayehu et al. [48] further validated the influence of rising temperatures under SSP scenarios on crop and land suitability, a trend clearly reflected in Gardulla’s future projections.

Furthermore, Wanyama et al., (2024) [5] cautioned that, while future scenarios may expand irrigation-suitable land in Uganda, they also pose risks of reduced water availability, a paradox that resonates with Gardulla’s SSP5-8.5 scenario, which, despite increased suitability, may face higher evapotranspiration and water stress. In the Abay subbasin, it is anticipated that annual rainfall will drop by 64.7 mm in 2050 and 86.3 mm in the 2080s, while temperatures will rise by 1.8 °C and 2.5 °C during the same periods [9]. According to Ademe [46], there will likely be a 4.2% to 16% drop in rainfall in the Ethiopian Rift Valley. In contrast, the RCP4.5 scenario, an earlier version of SSP4.5, predicts that the temperature will rise by 0.7% to 1.25% by 2050. These projections support the conclusions of the current study regarding the influence of climate change on surface irrigation suitability.

While the Gardulla study confirms broader regional trends, it distinguishes itself through the use of high-resolution Sentinel-2A imagery and localized climate downscaling via the Statistical DownScaling Model (SDSM), enhancing the precision of future projections. This methodological advancement supports more targeted irrigation planning and contributes to climate-resilient agricultural development. However, unlike studies that incorporate socio-economic and institutional dimensions, this research focuses primarily on biophysical parameters. As Alemayehu et al. [48] noted, technical suitability does not guarantee implementation without supportive governance and infrastructure. Furthermore, the assumption of static land cover and soil conditions in future scenarios may oversimplify dynamic environmental realities. These limitations highlight the need for future studies to integrate hydrological modeling, socio-economic data, and institutional capacity assessments to ensure holistic irrigation planning.

5. Conclusions

This study demonstrates that the suitability of land for irrigation is highly sensitive to climate change. Under current conditions, only a small fraction of the study area (8.5%) is highly suitable, with the majority classified as moderately or marginally suitable. Future climate projections under high-emission scenarios (SSP8.5) indicate a marked increase in highly and moderately suitable lands, reaching 14.2% and 62.0% by the 2080s, respectively. The northwestern zone is projected to emerge as a new hotspot for irrigation, while areas that are currently marginally suitable or unsuitable are expected to decrease. Climate projections under high-emission scenarios indicate an expansion of the highly and moderately suitable lands by the 2050s and more by the 2080s, particularly in the northwestern zone. In contrast, marginally suitable and unsuitable areas decline. The largest changes occur under high-emission scenarios, underscoring the strong sensitivity of irrigation potential to climatic shifts.

Therefore, this study indicates that irrigation potential is highly climate-sensitive and shifts dynamically with future climate scenarios, and it shows the benefits of integrating climate projections with land suitability analysis for better irrigation planning and adaptation methodologically.

Theoretically, these findings highlight the importance of incorporating climate-driven variability into assessments of irrigation potential, illustrating the dynamic interplay between environmental conditions and agricultural suitability. Methodologically, this study underscores the value of combining spatial analysis with climate projections to identify future hotspots and vulnerabilities, providing a replicable approach for assessing irrigation potential under changing climates. These insights can inform sustainable water management strategies and guide policy decisions for climate-adaptive agricultural planning.

6. Limitation and Recommendation

Despite these encouraging projections, limitations include reliance on static biophysical datasets, uncertainties in climate models, and the omission of governance and water availability factors. Soil property data from ISRIC at 250 m resolution provide useful regional coverage, but it may not capture field-level variability in texture, infiltration, and drainage that strongly influence surface irrigation performance. This generalization can reduce the accuracy of suitability assessments, highlighting the need for higher-resolution surveys or ground validation at the local scale.

Therefore, future research is recommended to integrate dynamic hydrological modeling, higher-resolution soil data, ground validation, and institutional variables to improve applicability. The findings suggest actionable evidence for policymakers, land-use planners, and stakeholders to design adaptive, climate-resilient irrigation strategies, such as adopting efficient water-saving technologies like drip and precision irrigation and using drought-tolerant crops, positioning the Gardulla Zone as a model for similar semi-arid and sub-humid regions facing climate pressures.