**1. Introduction**

For thousands of years, man has been coping with salinization processes in irrigated agriculture [1], which is the main consumer of water worldwide, accounting for nearly 70% of the total global water withdrawal [2]. This problem continues to worsen, and today, 25–30% of the world's irrigated lands in more than 100 countries are affected by salt [3,4]. Population growth, which increases the demand for both food and freshwater for domestic use, further contributes to this growing challenge, as it incentivizes the expansion of irrigated lands and the use of non-freshwater sources such as brackish water and treated wastewater (TWW) for irrigation [5]. These processes are further augmented by climate change, which increases irrigation needs due to higher vapor–pressure deficits [6] and reduces the natural enrichment of freshwater sources [7]. Moreover, the common agronomic solution to salinization is to apply excess amounts of irrigation water to leach the salts below the root zone [8]. However, that method gradually increases the salinity of groundwater bodies [9] and consequently counteracts its original purpose. The use of desalination, which is a remedy for both the growing water shortage and salinization, is steadily increasing [10], but it consumes a great deal of energy and entails high brine disposal costs [11].

The processes described above reflect a strong linkage between agricultural irrigation and the supply of water to different users—a link that should be accounted for in the design of sustainable and economically viable solutions to the problems of water shortages and salinization. This study focuses on irrigation practices in agricultural regions with access to

**Citation:** Slater, Y.; Reznik, A.; Finkelshtain, I.; Kan, I. Blending Irrigation Water Sources with Different Salinities and the Economic Damage of Salinity: The Case of Israel. *Water* **2022**, *14*, 917. https:// doi.org/10.3390/w14060917

Academic Editor: Guido D'Urso

Received: 10 February 2022 Accepted: 14 March 2022 Published: 15 March 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

several water sources of different salinities and evaluates the impact of these practices on water management in a multiregional water distribution network. Specifically, we evaluate the economic damage caused by salinity under different strategies for blending irrigation water sources with different salinities. We focus on the case of Israel—a country equipped with a complex water supply system and a large agricultural reliance on non-freshwater sources such as TWW and brackish water, which together, constitute about 60% of the country's total irrigation water. Our economic analysis accounts for the impact of blending options on agricultural cropping patterns, optimal long-term management policies, and the development of the Israeli water supply system.

Despite the vast agronomic literature on the production impacts of irrigation water mixtures (e.g., recent agronomic studies explore the impact of conjunctive use of water resources [12,13], employ GIS for assessing salinity impacts under different irrigation practices [14], and measure the impacts of water irrigation strategies on soil and plant properties [15]), economic analyses of that issue are scarce. Parkinson et al. [16] were probably the first to economically evaluate water blending options. Knapp and Dinar [17], Dinar et al. [18], Kan et al. [19], and Kan [20] employed field-level models to study the profitability of mixing water sources with different salinities for the irrigation of specific crops. Feinerman and Yaron [21] and Kan and Rapaport-Rom [22] incorporated blending options in regional-scale models, in which the land allocation across crops was endogenous. However, all of these studies assumed exogenous water supplies and therefore overlooked the implications of water management strategies within agricultural regions on the water economy as a whole. The contribution of this paper is the introduction of the nexus between the agricultural and water sectors into the economic analysis of water blending strategies.

The linkage between the intraregional management of irrigation water and the design of economy-wide water-supply systems is of particular importance in water economies that supply water to different users from multiple sources and/or where the recycling of domestic TWW in irrigated agriculture creates a strong interdependence between the two sectors. In such water economies, the optimal allocation of water across users depends on their demands for the various water sources, where the demand of any farming region for different water types depends on the irrigation practices in that region. This is the case in Israel, where the water distribution network connects almost all users and sources in the country. That connectivity implies that water usage at a particular place and time may have opportunity costs, as it cannot be used for other purposes at alternative locations and times [23].

Hydroeconomic models provide a powerful tool to analyze water management problems on different scales and under various spatiotemporal conditions (see [24–34]). However, to the best of our knowledge, the only hydroeconomic model that incorporates salinity considerations in the allocation of water to urban and agricultural users is the MYWAS-VALUE (Multi Year Water Allocation System-Vegetative Agricultural Land Use Economics) model, which was developed by Slater et al. [35] for the case of Israel. Slater et al. [35] employed MYWAS-VALUE to evaluate the societal deadweight loss entailed by overlooking the impact of salinity on agricultural production in the design of water infrastructures. However, the model presumes regional irrigation water blending; that is, all of the inflows of water sources into any agricultural region are mixed before they are applied to the irrigated crops. This assumption has two drawbacks: first, compared to field-level blending, regional water blending may increase the detrimental impact of salinity on agricultural production because it affects all of the irrigated crops in any given region, including both salinity-tolerant and salinity-sensitive ones. Consequently, the exaggerated salinity damage may motivate the faster-than-optimal expansion of desalination capacities. Second, it turns out that farmers in Israel rarely blend irrigation water from different sources (personal communication; Anat Levingert, Senior Manager of the Consulting and Professional Service of the Israeli Ministry of Agriculture (Shaham)). Thus, designing the long-term development of water infrastructures under the assumption of the regional blending of irrigation water

sources may yield results that both inflate the agricultural damage incurred due to salinity and that are inconsistent with reality.

In this paper, we analyze two irrigation water mixing scenarios: field blending (FB) and regional blending (RB). The difference between the scenarios with respect to the intraregional water supply system is illustrated in Figure 1 for a hypothetical region, in which farmers grow five crops and have access to three water sources with different salinities: freshwater, TWW, and brackish water. Under FB, farmers can select a specific combination of the three sources for each crop, whereas the RB scenario implies one combination for all crops. Note that, while both scenarios do not preclude the non-blending option, avoiding blending in the RB case implies that only one water source is used in the entire region, whereas the FB scenario enables farmers to use all of the water sources that are available to them by assigning a single water source to each crop.

**Figure 1.** Schematic illustration of the field and regional blending scenarios in an agricultural region with five crops that can be irrigated by three water sources with different salinity levels: freshwater, treated wastewater, and brackish water.

Our analysis is based on the MYWAS-VALUE framework. We first calibrate the model under the FB assumption to reproduce the observed situation in a baseline year (2019). Then, we run the model under the FB and RB scenarios for a period of 30 years. We found that switching from FB to RB slightly expedites the development of desalination plants, but the average irrigation water salinity increased due to the reallocation of water sources across sectors and crops. Although salinity-sensitive crops face the largest reduction in per hectare production, the combined impact of changes in the (endogenously determined) prices of irrigation water and agricultural outputs motivates farmers to shift more water and land to the production of these crops.

We consider three measures of the economic damage caused by salinity under the two blending scenarios. The first measures the reductions in the agricultural production value caused by the design of water infrastructures that ignore the impact of irrigation-water salinity on agricultural production. This reduction amounts to USD 1195 and USD 1326 per hectare under the FB and RB scenarios, respectively (all monetary values are in terms of the 15th year of the 30-year planning horizon). The second measure is based on the negative relationship between the irrigation water value of the marginal product (VMP) in an agricultural region and the average salinity of the region's irrigation water; on average, the VMP decreased with the salinity by USD 0.39 and USD 0.30 per dS m−<sup>1</sup> per cubic meter of irrigation water for the case of FB and RB, respectively, or by −2.4 and −1.6 in terms of elasticity (note that both VMP and salinity are endogenous in MYWAS-VALUE). The last measure computes the marginal damage caused by salinity based on the shadow

values of the salt-balance constraints along the water delivery system; a salinity increase of 1 dS m−<sup>1</sup> costs USD 525 and USD 534 million a year for the whole country under FB and RB, respectively. Per cubic meter of irrigation water, we achieved USD 0.42 per dS m−1, with minor differences between the blending scenarios being observed.

The following section briefly describes the MYWAS-VALUE model; Section 3 compares the results under the two blending scenarios and discusses the three measures of the economic damage of irrigation water salinity; Section 4 concludes the paper, and Section 5 discusses the limitations of the analysis and avenues for future research.
