Investigation of managed aquifer recharge site suitability through multi-tiered decision making approach

Abstract

Globally, groundwater resources encounter notable challenges encompassing issues of chemical contamination, depletion, anthropogenic activities, and seawater intrusion. These challenges are further exacerbated by the prevailing arid climatic conditions and rapid population growth, resulting in restricted recharge and substantial overdraft, respectively. Consequently, the implementation of sustainable measures becomes imperative to facilitate the efficacious and enduring stewardship of these invaluable resources. While diverse management strategies have been proposed and adopted worldwide over the years to contend with these threats, ranging from optimizing pumping rates to enhancing water use efficiency, managed aquifer recharge (MAR) stands out today as one of the most promising technological approaches for addressing the degradation of subsurface resources. MAR entails the injection of reclaimed water (such as stormwater and treated wastewater) into the aquifer via various recharge structures, either for subsequent retrieval during dry period or for environmental enhancement purposes. Besides their acknowledged efficacy in replenishing stressed aquifers, MAR technologies are highly susceptible to instances of failure or suboptimal performance if not meticulously executed within appropriate sites. The process of identifying suitable locations for the construction of recharge facilities is notably intricate, owing to the multitude of decision parameters at play. This challenge is further compounded in countries where MAR remains a novice subject, thereby amplifying the complexity of the task. The exploration of MAR potential through the utilization of integrated geographic information system and multi-criteria decision analysis (GIS-MCDA) techniques is acknowledged within the research community as a valuable and insightful approach. Nonetheless, the MCDA-based decision frameworks proposed in the pertinent literature often lack robust mathematical underpinnings. Furthermore, while the mapping of MAR site suitability is commonly conducted at the watershed level, its implementation at the national and urban scales tends to be overlooked. Hence, besides providing an in-depth literature survey about the state of the art in GIS-MCDA based MAR sites suitability mapping, the studies encompassed within this thesis introduce novel and holistic contributions to the scholarly discourse, spanning both conceptual and methodological dimensions. Within the scope of this thesis, robust and reliable multi-tiered decision making frameworks designed to evaluate the potential of MAR across diverse geographical scales, encompassing national, watershed, and city extend were proposed. Moreover, we introduced a coupled groundwater vulnerability and MAR potential mapping model for the sack of identifying promising regions with minimized vulnerability to contamination for MAR implementation. In the first study conducted in this thesis, a robust approach that combines fuzzy analytical hierarchy process (AHP) and technique for order of preference by similarity to ideal solution (TOPSIS) to delineate suitable sites for MAR structures at a countrywide scale was proposed. The proposed model is applied to Djibouti, a hot, dry, and water-stressed country. A set of nine decision criteria were identified and a pairwise comparison survey was conducted to determine their relative importance. Additionally, the TOPSIS method was employed to integrate the decision layers and prioritize the study area. The results highlight the significance of rainfall, slope, and NDVI as the most influential decision parameters, while drainage density has the least impact. The suitability analysis reveals that 10.63%, 23.20%, and 31.06% of the country have very high, high, and moderate potential for MAR activities, respectively. Furthermore, an evaluation of viable MAR technologies within the nation is conducted based on the specific conditions prevailing in Djibouti. At the watershed level, the second study within this thesis focuses on a significant MAR technology extensively employed in arid and semi-arid regions: rainwater harvesting (RWH). RWH offers a promising strategy to address the pervasive challenge of water scarcity in such regions. It involves capturing stormwater for domestic and agricultural use while contributing to aquifer replenishment. In this research, a hybrid approach is proposed, integrating fuzzy AHP and VlseKriterijumska Optimizacija I Kompromisno Resenje (VIKOR) algorithms. This approach is implemented in the water-stressed southeastern basin of Djibouti. This study represents the pioneering endeavor to comprehensively assess RWH potential in Djibouti. Despite existing RWH practices in the country, their implementation often lacks robust scientific foundations, resulting in reduced effectiveness and operational vulnerabilities. A hierarchical structure with 12 decision parameters is formulated, organized into 3 dimensions: physical, environmental, and socio-economic. These dimensions are further divided into 4 specific criteria, establishing a comprehensive framework for addressing the complex issue. Notably, this research extends the incorporation of socio-economic parameters beyond constraint mapping, encompassing suitability analysis as well. Additionally, the application of the VIKOR algorithm for RWH mapping is unprecedented in the literature, marking a pioneering contribution. Findings highlight the significance of the environmental (47.67%) and physical (37.98%) dimensions, with the socio-economic cluster (14.35%) exhibiting lower importance in the fuzzy AHP analysis. Also, the suitability map reveals a substantial area, approximately 872.5 km2, with moderate to very high suitability for various RWH structures like check dams, percolation tanks, and terracing. Regarding the third study, the potential of drywells in an urban environment was investigated using multi-tiered decision approach. In fact, despite existing several MAR technologies, land availability for installation and limited municipal budget restrict their usage, especially in urbanized areas. Consequently, drywells emerge as a highly regarded solution for replenishing stressed shallow urban aquifers, primarily owing their reputation to their minimal land requirement and cost-effectiveness aligned with municipal budgets. Nevertheless, the widespread implementation of drywells is hampered by the absence of comprehensive guidelines and frameworks governing their sitting. In this third research endeavor, a coupled fuzzy AHP and TOPSIS model is proposed to identify favorable zones for drywell construction in the urban city of Kayseri, Turkey—an area prone to declining water tables. It is noteworthy that this research represents the initial attempt in investigating the site suitability of drywells through an integrated GIS-MCDA technique. Based on an in-depth literature survey and technical reports relevant decision criteria were selected and their relative weights were determined through the fuzzy AHP. The results indicated that hydraulic conductivity, depth to groundwater, soil texture, and land use/land cover were the most determining criteria with 33.9%, 17.8%, 11.2%, and 10.1% of weights, respectively. The validation of these criterion weights was conducted by comparing them with those from previous MAR suitability mapping efforts in relevant literature. Additionally, a sensitivity analysis was performed to underscore the robustness and stability of the findings under various degrees of fuzziness. Finally, the study area was divided into 54 sub-regions based on the road-network of the city and the TOPSIS was applied to prioritize the sub-regions to help identify promising areas for the vadose zone wells construction. Concerning the fourth study in this thesis, a combined groundwater vulnerability and MAR sites suitability mapping model was introduced in the pertinent literature. Notably, while drywells stand out as a superior MAR technology in urban environments, the task of identifying appropriate locations with minimal susceptibility to contamination for drywell projects presents an intricate challenge. Hence, the present research proposes a framework encapsulating the combined assessment of groundwater vulnerability and MAR sites suitability analysis to pinpoint the most featured areas for installing drywells in Kayseri, Turkey. To extrapolate the vulnerable zones, not only the original DRASTIC but also its MCDA based modified variants were evaluated with regard to different hydrochemical parameters using the area under the receiver operating characteristic (ROC) curve (AUC). Besides, the fuzzy AHP rationale was adopted to signify the importance level of criteria and the robustness of the framework was highlighted with sensitivity analysis. In addition, the decision layers and the attained vulnerability layer were combined using the weighted overlay (WOA). The findings indicate that the DRASTIC-SWARA correlates well with the arsenic (AUC=0.856) and chloride (AUC=0.648) and was adopted as the vulnerability model. Groundwater quality parameters such as chloride and sodium adsorption ratio, as well as the vadose zone thickness, were found to be the most significant decision parameters with importance levels of 16.75%, 14.51%, and 15.73%, respectively. Overall, 28.24% of the study area was unsuitable for recharge activities with high to very high vulnerability, while the remaining part was further prioritized into low to high suitability classes for MAR application. The proposed framework offers valuable tool to decision makers for the delineation of favorable MAR sites with minimized susceptibility to contamination. The decision frameworks proposed within this thesis empower policy-makers and stakeholders to efficiently discern and prioritize regions exhibiting promising MAR potential across various geographical scales. Particularly, the outcomes from the initial two investigations in this thesis hold primary relevance for Djibouti's water resources authorities, aimed at bolstering sustainable water resource management through MAR and RWH strategies in the country. Regarding the third study, it is believed that the proposed GIS-based hybrid MCDA framework in delineating sites favorable for drywells construction can be adopted by municipalities lacking guidelines and standards for sitting drywells. Furthermore, it is anticipated that the integrated model presented in the final study of this thesis can facilitate the sustainable management of aquifers and contribute to the mitigation of water scarcity challenges in various global regions, all while minimizing the risk of MAR project failure due to the installation of recharge facilities in areas susceptible to groundwater contamination.