Introduction

There is a broad range of purposes for managed aquifer recharge (MAR) as illustrated in this Special Issue. Papers here address: water quality improvement for drinking water supply in Finland and India; enhancement of groundwater storage for irrigation supplies in India; restoration of depleted aquifers and security against drought for urban water supplies in the USA, Portugal and Spain; preservation of groundwater dependent ecosystems stressed by irrigation in NZ and USA; flood water harvesting for productive use in India and Mexico; storage of recycled wastewater for irrigation in Mexico, Spain and Portugal, and for drinking water supplies in Namibia. In addition to a range of purposes, there is also a range of types of MAR described in this Special Issue; streambed recharge structures, river bank and esker filtration, surface water spreading, and recharge wells. Types of water used for recharge include lake and river water, recycled water and urban stormwater. The types of aquifers used for storage or treatment in this issue cover alluvial, fluvio-glacial, fractured rock and karstic, and include unconfined and confined aquifers. Table 1 summarises these attributes for the MAR projects and investigations reported in this Special Issue. These follow the order in which papers appear, starting with multi-site assessments, descriptions of individual operating projects, trials and pilot projects, and preliminary investigations.

Table 1 Summary of papers in special issue in relation to stage of development of project, purpose, method, source water, and location, with a keyword descriptor of main focus
Full size table

Planning and implementing MAR projects

Synthesis from multiple sites

In every case these papers describe MAR projects that involve integrated management of surface water and groundwater resources, where excess surface water of some type is intentionally stored in aquifers either for water quality improvement before use, for increasing or sustaining supplies during dry periods, or for sustaining groundwater-dependent ecosystems. Although there is a diversity of projects, they together affirm some basic messages that will be valuable to those contemplating MAR for the first time.

First, any new MAR project developer is not alone. There are a host of projects undertaken around the world, with at least 1200 so far reported by Stefan and Ansems (2018) from 62 countries across all inhabited continents, as part of the outcomes of an IAH-MAR Global MAR Inventory working group. Information on each project is available on the MAR Portal (IGRAC 2018) that can be searched for attributes (purpose, water type, aquifer type, method, country) relevant to an intended project, to reveal existing experience. Considerably more information is available for MAR projects in Latin America and the Caribbean (Bonilla Valverde et al. 2018) that are also reported on the IGRAC MAR Portal. Among the 340 Mm3/year MAR reported almost half was in Mexico. Novel examples described were more than 50,000 small recharge structures in north east Brazil, stormwater injection wells in Cuba and injection for subsidence control in Mexico City. Increasing the information on projects on the MAR Portal would be most valuable particularly for attribute combinations that are currently missing or low in number, or in countries sparsely represented. Operators of such projects are encouraged to upload basic information on the brief proforma found on the MAR Portal.

Information on the economics of MAR on a site specific basis has been difficult to find in the past and Ross and Hasnain (2018), as part of an IAH-MAR working group on Economics of MAR, have assembled information on costs for 21 sites for which data were available. An important part of Ross and Hasnain’s contribution is to lay out a pragmatic framework to allow consistent evaluation of costs appropriate for water supply or water banking types of projects. Several other papers as shown in Table 2 also provide cost information for specific projects. As stated by Ross and Hasnain, further work is needed to establish the benefit–cost ratio for MAR projects, and comparisons with regard to alternative water management projects that have equivalent impacts would be useful.

Table 2 Contributions of the papers in this special issue with respect to information perceived to advantage the uptake of MAR
Full size table

San-Sebastian-Sauto et al. (2018) set out on an ambitious task to characterise and benchmark seven varied MAR projects in Portugal and Spain. Two Spanish projects had data for 13 years and one for 6 years and the four Portuguese sites were up to 2 years old. The attributes of each of the seven sites are systematically documented to give a description of their design, volumes recharged, means of operation, water quality status, capital and operating costs and energy use. Due to very large differences between site operations it was difficult to find metrics for meaningful benchmarking of the relative performance of sites. However one parameter, mean annual infiltration rate, was found to be a very valuable diagnostic at each site on rate of clogging and need for remediation. Low capital costs and low energy consumption were found, which would help favour MAR, but no alternatives to MAR were considered in benchmarking.

Reflections on operational experience

The next four papers describe four operational sites each of which has more than ten years history. Murray et al. (2018) describe the investigations and progress made in expanding a MAR scheme to secure drinking water supplies for Windhoek, the capital of Namibia. This uses a combination of natural water and advance-treated recycled water before injecting about 0.5 Mm3/year in complex fractured quartzite aquifers. This case was shown to be much more economic than alternative supplies and novel institutional arrangements were established to support sustainable operations. This project has been operating effectively since 2006.

Tanttu and Jokela (2018) describe the use of fluvio-glacial deposits to remove natural organic matter and reduce temperature variations of lake water before supplying drinking water in Finland. One month of residence time in the aquifer is sufficient to improve the quality for use. Basins, sprinklers, and injection wells are compared as alternative recharge systems. The issue of managing enriched levels of iron and manganese in recovered water is addressed, and demonstrates a sustainable, economic solution.

Riverbank filtration is a practical low-cost way to improve the quality of river water for drinking water supplies and Sandhu et al. (2018) report typically 3–4-log removal of E. coli at riverbank filtration sites in India. However, their paper focuses on a periodic operational problem of safety of water supplies from a riverbank filtration site adjacent a tributary of the River Ganges at Srinagar, Uttarakhand, India. Their work reveals that short-circuiting occurs in floods when bank filtration wells on the floodplain are submerged, in spite of the best efforts to prevent flood water entry. They suggest that shock chlorination is unlikely to be effective and that such wells need to be disconnected from the supply during flood.

The final paper in this set on operational MAR sites describes 10 years of experience in recharging reclaimed water via soil aquifer treatment at San Luis Rio Colorado in north-west Mexico. Humberto et al. (2018) record the water quality in ambient groundwater, the recycled water in the recharge pond and groundwater at various depths, and illustrate that the reported parameters are continuously within the regulatory requirements for Mexico and that the treatment rate is being sustained. They also demonstrate that limiting the total suspended solids concentrations in pond water and restricting the pond water depth to 0.3 m, together with periodic scraping or disking of basins has been effective in reducing the rate of clogging to a manageable level.

Implementation and evaluation of trials or pilot sites

The next set of five papers describe implementation and evaluation of trials or pilot sites. Lluria et al. (2018) describe the investigations that have enabled the locations for a mix of injection-only and aquifer storage and recovery wells to be determined that optimise the storage of surface water in a fractured granitic aquifer in central Arizona. Injection rates exceeded recovery rates due to fractures above the water table. This provides a good example of the value of characterising the aquifer well and being able to use that information for optimising the design of a system to recharge almost 4 Mm3/year.

The next two papers by Scherberg et al. (2018) and Painter (2018) describe schemes in temperate irrigated plains in Oregon USA and Canterbury NZ where seasonal groundwater-dependent streamflow is vital for aquatic ecosystems. In each location, irrigation with surface water and groundwater occurs. Irrigation canals are being lined to prevent leakage and reduce surface water diversions. However, this also reduces recharge to the alluvial aquifer and base flow. Hence infiltration basins are being introduced in strategic locations to boost environmental flows to groundwater-dependent streams. The aim in both cases is to site MAR structures to allow optimum conjunctive use of groundwater and surface water for irrigation while still protecting ecosystems. Scherberg et al. (2018) focuses on modelling of the surface water and groundwater system to optimise the scale and locations for MAR, based on expanding an existing 11 Mm3/year scheme. Painter (2018) also uses a model but the thrust of the paper is on understanding the water quality requirements and engaging with the community to more clearly define the project objectives. Initial field testing results suggest MAR could be an effective way to meet the groundwater-dependent ecosystem requirements.

The next pair of papers are set in a semi-arid area of Rajasthan where check dams have been constructed in ephemeral stream beds for more than 30 years to harvest monsoon runoff allowing time for water to infiltrate and replenish a hardrock aquifer that supports irrigated agriculture. Dashora et al. (2018) evaluates recharge from four check dams over 2 years using a simple water balance method based on farmer measurements of daily water levels. Dashora et al. (2018) also reviewed published studies of almost 30 check dams and compared results in relation to hydrological variables. Jadeja et al. (2018) explored participatory approaches to groundwater management in India, explaining the capacity building program with farmers that enabled Dashora et al.’s work, and also provided rainfall data and detailed information on groundwater levels. That project, under the title managed aquifer recharge and village-level interventions (MARVI) (Maheshwari et al. 2014) helped farmers to interpret and share the information with the local village communities and form groundwater cooperatives to manage their use of groundwater, and assist in ensuring maintenance of streambed recharge structures, and improve linkages with catchment and water resources management programs.

Approaches to planning and preliminary studies

The final set of five papers involve planning and preliminary studies performed to select MAR sites, to design pilot projects or to explore the feasibility of extensive MAR programs. The first of these, by Chinnasamy et al. (2018), describes linking of hydrologic, groundwater and flood inundation models to estimate the potential impact of widespread recharge of floodwaters in the ~ 19,000 km2 Ramganga river basin, a sub-basin of the River Ganges, India. The modelling revealed that if implemented, this could have significant benefits in restoring a depleted aquifer, would substantially mitigate peak flood flows and would increase base flows in the river. A field demonstration project at local level has subsequently commenced to enable ground-truthing and subsequent scale up if successful.

At a contrasting scale, Mäkinen et al. (2018) present the results of a field study over a 30 km2 area of Finland where detailed characterisation of the esker depositional environment was required to refine the design of an aquifer storage transfer and recovery project (that is, with separated injection and recovery wells) in glacio-fluvial deposits to improve water quality [as discussed by Tanttu and Jokela (2018)]. Observation wells, geophysics, remote sensed temperature data for groundwater springs, and a 9 months tracer test were used to support refined groundwater modelling. This revealed a bifurcating flow path between injection and recovery with bi-modal travel times. As a result, the recharge site was relocated to produce a simpler to manage system.

Wurl and Imaz‑Lamadrid (2018) address the design of a runoff harvesting system consisting of four dams upstream on different rivers with releases into stream beds and infiltration dams to augment recharge by more than 20 Mm3/year to help sustain levels and water quality in a 724 km2 irrigation area in Baja California, western Mexico. Initially, the paper focuses on the model and its calibration, using data from soil texture analysis from 554 farms. Next, 8 sites were selected for hydraulic testing by drilling and coring for hydraulic tests leading to selection of four sites for recharge. The groundwater model, calibrated on monitored groundwater levels and known rates of groundwater extraction revealed that the recharge operation would help sustain the central Santo Domingo Valley and would slow but not prevent saline coastal saline intrusion due to excessive groundwater extraction.

The last two papers involve the application of an unsaturated porous media model as part of the design of pilot scale infiltration basins. Palma Nava et al. (2018) described the preliminary investigations required to establish a trial for recharge with recycled water via soil aquifer treatment to help sustain the depleting groundwater supply for the city of Chihuahua in northern Mexico. The site characterisation depends on surface geophysics using vertical electrical soundings and the lithology derived from a well. This was used in an unsaturated flow model (VS2DTI) to predict the results of a future infiltration basin trial and hence to design the trial and its monitoring regime. Water quality analyses as required under Mexican guidelines for recharge of recycled water, when completed would enable either approval for the trial with monitoring or require further treatment before such approval is given. Sallwey et al. (2018) also undertook unsaturated zone modelling (with HYDRUS3D) to assist with the design of a very small scale infiltration test. This was valuable in determining the timing and spatial distribution of vertical water flux at a depth corresponding with the water table under different potential operating regimes. They also used the model to explore the best locations for field-monitoring equipment. Importantly, this paper also reviews a number of approaches to unsaturated zone modelling at MAR infiltration sites and probes the strengths and weaknesses of models for such applications.

Conclusions on planning and implementing MAR projects

Numerous MAR project implementers have reported on issues and information gaps that need to be addressed or would be helpful to know at various stages in establishing managed aquifer recharge projects. In this Special Issue, each paper, as discussed above, makes a contribution to the body of knowledge that helps fill these gaps, as shown in Table 2. This table may be used as a directory to help find information on issues most important to developers and regulators of future MAR projects.

Each paper has been briefly discussed above, and although these papers cover a wide range of situations and stages of implementation there are some common threads. Projects proceed from a concept to on-the-ground reality by a sequence of steps to validate the concept, select sites, allow practical demonstration to win community and regulator support, prove economic viability, and to learn how to monitor, manage and regulate projects to ensure they are sustainable. These steps may take considerable time, especially for projects that have great strategic value or have high perceived risk. In cases such as Windhoek (Murray et al. 2018), it has taken a decade to confidently produce crucial strategic reserves in complex geology. A key message is to start investigations well before the MAR project is actually needed to provide the necessary confidence that it will perform as required.

The importance of aquifer characterisation stands out in the completed projects. Wherever geology is complex, such as for heterogeneous aquifers, fractured hardrock aquifers, karstic aquifers, in or near fault zones, where aquitards are thin and variable, where basements are contorted, boundaries are unknown, stream–aquifer interaction occurs or aquifers contain water of variable or poor water quality, the attention to understanding groundwater dynamics needs to be considerably greater than for simpler settings. This may require multiple lines of evidence such as drilling and coring, geophysics, aquifer pumping tests, piezometers, hydrochemical, isotope and tracer studies and modelling to gain an adequate understanding of the storage or treatment zone. This will be needed to determine the capacity of the aquifer to accept water, the rate at which water may be stored, the fate of stored water and the ability to recover it in the short and long-term and the uncertainty associated with these estimates.

As scientific and technical methods and knowledge advance it is anticipated that the future key issues may shift. Institutional arrangements, and community engagement will become increasingly important for integrated water management and are touched on by several papers here. Governance arrangements for MAR such as entitlements to water for recharge, to storage space in aquifers and to recover water are also expected to become increasingly important as demand grows. Uptake of frameworks to do this (such as Ward and Dillon 2011) has been slow and not addressed in this special issue, but will be needed to accelerate to address cumulative impacts of proximal MAR projects, and to embed MAR as a tool in conjunctive management of water resources (Evans and Dillon 2018) as called for in the ISMAR9 Call to Action: Sustainable Groundwater Management Policy Directives (Parker and Villarreal 2016).