Assessment of using ponds as a nature-based solution on the optimal irrigation management of paddy fields

In many countries, water shortages for irrigation have caused significant water crises (Fang et al. 2012). This problem has become more challenging due to population growth and increasing water demands for food security (Wu et al. 2021). Proper management and optimal use of agricultural water are crucial for the sustainability of water resources, as irrigated agriculture is the largest water consumer in many regions. Rice, which is cultivated as flooding in Guilan Province, Iran, is one of the most water-intensive agricultural products (Hadizadeh et al. 2018).

Rice is not only a highly important grain worldwide but also holds great strategic value in Iran. As a staple food in numerous developing and Asian nations, the availability of water for rice cultivation has a critical impact on the food security of these countries. Since the absence of food security significantly affects political, cultural, and social aspects, addressing the irrigation requirements for rice has become an urgent concern for regional officials and planners (Ebrahimian et al. 2020).

To address water scarcity and reduce pressure on a single source, the conjunctive operation of multiple water sources, such as rivers, groundwater, and reservoir dams, can significantly improve water consumption efficiency and minimize losses due to evaporation and transpiration (Liu et al. 2013; Golmohammadi et al. 2017; Banihabib et al. 2018; Seo et al. 2018; Nayyeri et al. 2021). When a single water source may not be adequate in terms of quantity or quality, the conjunctive operation of existing water sources can offer a viable solution (Shi et al. 2012; Harmancioglu et al. 2013). Previous studies have utilized linear programming optimization models (Cheng et al. 2009) and the conjunctive operation method (Singh 2015) to optimize water and land resource management, resulting in improved agricultural outcomes. These studies have shown the benefits of using surface water and groundwater conjunctively, including increased agricultural income and reduced soil salinity (Das et al. 2015; Li et al. 2018). Multi-objective optimization models have also been developed for the effective allocation of surface water and groundwater resources (Naghdi et al. 2021).

During the operation of different water sources, priority is given to utilizing a source based on its natural characteristics to avoid losing a significant portion of available water sources. Ecosystems can store water by creating a storage space and considering the soil's water infiltration capacity (Uy & Shaw 2012). Ecosystem management and the use of nature-based solutions in water resource operation can mitigate drought effects, enhance groundwater recharge, improve water quality, and be cost-effective (Cohen-Shacham et al. 2016; Masi et al. 2017). Consequently, implementing nature-based solutions at various levels has proven effective in addressing environmental and socio-economic challenges and promoting business utilization (Faivre et al. 2017; Lafortezza et al. 2018; Aldieri et al. 2020).

Using the capacity of water storage volume in ponds to allocate water resources is a nature-based solution that not only creates a natural landscape for recreation and entertainment but also serves as a suitable water storage tank to prevent wastage of water resources (Magaudda et al. 2020; Staccione et al. 2021). Yoon (2009) demonstrated that ponds are a primary water supply source for paddy fields during drought conditions. In coastal areas like northern Iran, where rainfall is significant during the non-agricultural season, many rivers and streams are seasonal and can become inaccessible if not properly managed, ultimately resulting in water loss. In such cases, utilizing the stored water volume in ponds can alleviate water resource shortages and fulfill the water requirements during the cropping season (Ebrahimian et al. 2020).

Although there have been a few studies on using ponds for water resource allocation, they have mainly focused on increasing the storage capacity of ponds through dredging to optimize water resource allocation in paddy fields (Fang et al. 2012). Other studies have evaluated the effects of ponds on the water balance cycle, environment, and their contribution to water supply, considering the impact of climate change (Staccione et al. 2021). Ebrahimian et al. (2020) utilized the conjunctive operation of canals and ponds to optimize the allocation of available water resources in the Tajan-Iran irrigation network.

However, no study has investigated the increase in the storage volume of ponds by considering the hydraulic connection between pond water and the groundwater level of the adjacent aquifer. Therefore, this study aims to maximize the storage capacity of ponds by exploring the scenario of increasing their volume through dredging and deepening. In the pond deepening scenario, the excavation depth will be determined based on the hydraulic connection between the water ponds and the groundwater level. This approach has not been previously explored in water allocation planning studies, highlighting the novelty and significance of this research.

The Sefidroud irrigation and drainage network was implemented with the construction of the Sefidroud Reservoir Dam on the west coast of the Caspian Sea, located in Guilan Province. This network covers an area with an altitude ranging from −26 to +100 m and has a Mediterranean climate.

The average annual rainfall in this region is approximately 1,200 mm, with about 70% occurring during the months of October to March. The main crop in this area is rice, which occupies around 95% of the annual cultivated land. The Sefidroud River serves as the primary water source for agricultural activities in the Sefidroud irrigation and drainage network. The total area of this network is about 284,180 ha, consisting of three irrigation areas: Central (G), Foumanat (F), and East (D).

The G irrigation area is connected to the left bank of the Sefidroud River from the east, the Foumanat area from the west, the sandy shores of the Caspian Sea and the Anzali wetland from the north, and the Geleroud Dam, a part of the Rasht-Tehran Road, and the left channel of Sangar from the south. The geographical area of G is 118,250 ha. This irrigation area is divided into seven command areas: G1 to G7. G1, G5, G6, and G7 have a modern network, while G2, G3, and G4 have a traditional rivers network. The water supply for this area comes from the regulated flow of the Sefidroud River, which is directed through the left side channel of the Sangar diversion dam with a capacity of 114 m³ /s. The rice cultivation area in G is 78,502 ha, accounting for 66.5% of the total area of G and 96% of the overall cultivated land in this area.

The study area's water resources are categorized into surface water and groundwater to meet various water demands, including domestic, industrial, agricultural, and environmental needs. Surface water data encompass water released from the Sefidroud Reservoir Dam, local rivers, and ponds. Groundwater sources are accessed through pumping wells in the study area.

The water released from the Sefidroud Dam is distributed throughout the Sefidroud irrigation and drainage network using channels. This distribution occurs through the Tarik diversion dam, which divides the water into two parts. One part flows into the Foumanat area through the Fouman water tunnel, while the other part reaches the central and eastern regions after passing through the Gelehroud diversion dam. After traveling hundreds of kilometers through the main and sub-channels, the right and left channels of the Sangar Dam irrigate the agricultural lands in the command area before eventually flowing into the sea. G3, the fourth and final command area, shares a border with the Sefidroud River. The traditional Noroud River, along with its main tributaries including Armajub, Gafshe-Roud, and Khoshke-Bijar, play a crucial role in the transfer and distribution of water within the G3 command area. The water released from the Sefidroud Dam enters the Noroud Canal at the Sangar diversion dam, and subsequently, it is distributed through a traditional stream as illustrated in Figure 1.

Rivers have been crucial water sources for human societies throughout history. Even today, rivers continue to play a vital role. In Guilan Province, the abundance of springs, high groundwater levels, and the proximity of mountains to the study areas and plains have resulted in the formation of numerous rivers. However, in the G3 command area, water is not sourced directly from local rivers. Instead, the existing rivers traditionally distribute the water received from the canal.

Ponds serve as an additional source of surface water. They can occur naturally or be created through human intervention. In the northern region of the country, ponds play a crucial role in supporting the ecosystem of migratory and tourist birds, in addition to meeting agricultural and fishing water demands. Ponds also serve as important water storage facilities during droughts, specifically for irrigating paddy fields. During non-agricultural seasons, rivers and runoff resulting from rainfall flow into the ponds, which are then utilized for irrigation purposes through pumping and gravity during the spring and summer seasons.

Figure 1 illustrates the location of ponds in the G3 command area. This area comprises 31 ponds, covering a command area of 972.5 ha. The depth of the ponds ranges from 1 to 1.8 m, with an average depth of 1 m considered for this study. The reservoir volume of the ponds is calculated based on this average depth. It is worth noting that the presence of water-loving plants in the ponds leads to an available water factor of 50%.

Groundwater sources in Guilan Province are replenished by inflows from mountains, infiltration of rainfall, surface flows, and the return flow from domestic and industrial usage. The province's high seasonal rainfall, abundance of urban rivers, and elevated groundwater levels have resulted in the formation of aquifers with significant water storage capacity. Despite the favorable conditions, farmers in the studied area have not extensively utilized groundwater resources due to the high cost of pumping. Their priority remains meeting the irrigation needs of paddy fields. However, excessive extraction of groundwater in certain areas, without considering the aquifers' capacity, has resulted in the depletion of wells and water scarcity in numerous springs.

Water consumption in the study area is categorized into domestic, industrial, agricultural, and environmental usage, with agricultural consumption being the most significant. The total area of the G3 command area is 23,611.7 ha, with approximately 72% dedicated to paddy fields, 14% to mulberry gardens, and the remainder to other crops. Since mulberry gardens and other crops rely on rainfall, this study focuses solely on the water requirement of rice, as it is the only crop that requires irrigation. The estimation of crop evapotranspiration was conducted using a soil water balance-based model, which will be explained in detail in the methodology section.

The primary factor in calculating plant water demand is plant evapotranspiration (Cohen-Shacham et al. 2016). This can be achieved through irrigation planning models that utilize simulations of soil water balance. Typically, these models calculate ETc by multiplying a vegetation coefficient (K_c) with reference evapotranspiration (Allouche et al. 2014). One such model is the SIMDualKc model, which employs the dual vegetation coefficient method to estimate ETc (Raes 1982; Allen et al. 1998, 2005, 2007). This model was specifically developed for irrigation planning across various vegetation types, including garden crops and partial cover crops like vegetables. Moreover, the model incorporates optimization of irrigation strategies for supplementary irrigation management, which involves the simultaneous use of rainfall and irrigation water. The model utilizes the Penman–Mantith method by ASCE-EWRI (2005) in a summarized form to standardize the calculation of grass reference evapotranspiration.

⁠, ⁠, and are the decision variables of the optimization model, which are the amount of water allocated through the canal, groundwater, and ponds, respectively (million m³). is the irrigation requirement (million m³); is the net irrigation requirement (m³/ha); A is the cultivated area (ha); is the irrigation efficiency; t, time; ⁠, ⁠, and are the amount of available water through the canal, groundwater, and ponds, respectively (million m³). m is to normalize the values of the objective function to be in a smaller range, which in this study m equals 100.

To maximize the potential of the ponds, the study explores the scenario of improving them through dredging and deepening operations. Dredging and deepening existing ponds is a common method used to increase their volume. However, in Guilan Province, where the groundwater level is close to the surface, it is important to ensure that deepening the ponds does not lead to water draining from the aquifer into the ponds. It is obvious that the extraction of groundwater is expensive and increasing the depth of the ponds will allow the groundwater to enter the ponds without additional cost. Farmers can then use this water for irrigation as needed. But this method is not suitable in areas such as Guilan, where significant evaporation from the water surface occurs during the crop growing season. It is worth noting that creating these conditions may not be ideal in terms of water quality, as water entering the pond from the aquifer may be more polluted than the water in the aquifer itself. Therefore, it is possible to increase the depth to the point where the groundwater does not enter the ponds.

During the rice cultivation season, the climate conditions in the region result in significant evaporation from the ponds, causing a substantial amount of groundwater, a valuable water resource, to be lost through evaporation from the water surface. Based on the daily pan evaporation data in the study area and using the pan evaporation method, the evaporation volume in May, June, and July, for the existing ponds, was found to be 0.028, 0.04, and 0.045 million m³, respectively. Based on the results obtained from the scenario of increasing the depth of ponds, the evaporation volume values were 0.54, 0.76, and 0.45% of the available water volume of ponds in the months of May, June, and July, respectively. Therefore, the method of deepening the ponds can be considered, but careful consideration should be given to the groundwater level.

To determine the amount of soil removal required for each pond in the G3 area, the Thiessen polygon method was used based on the number and distribution of observation wells in the entire G3 command area. The effect zone of each well was determined in terms of the pond surface it covers. Subsequently, the ponds corresponding to each well were analyzed in terms of the required drilling depth based on the groundwater level.

In the previous section, it was discussed that the G3 command area relies on canals, groundwater, and ponds as water sources. To determine the maximum allowable groundwater withdrawal, the MODFLOW groundwater model was utilized. By implementing the MODFLOW model and analyzing the monthly groundwater balance under current conditions, the maximum monthly withdrawal values were determined for equilibrium conditions where the groundwater balance is zero.

Table 3 provides a summary of the permitted withdrawal values for each water source, expressed in million m³. It also includes the available water amounts from canals and ponds. It is important to note that Table 1 presents the monthly irrigation requirements for rice, measured in million m³.

The maximum allowable depth of deepening and the corresponding increase in pond volume are presented in Table 5. The optimization model was then executed for the conditions outlined in Table 5, serving as the first scenario. Subsequently, the first scenario was analyzed with a 20% increase in efficiency, creating the second scenario. Earth level and groundwater level in Table 4 refer to the level of the land surface and the groundwater level, respectively, relative to the free sea level.

The current state of groundwater and pond depths was assessed. It was found that out of the 31 ponds in the G3 area, only 8 ponds have a bottom level above the average groundwater level, while the remaining ponds are below the groundwater level. Ponds with bottom levels below the groundwater level cannot be deepened. The available water volume values were determined for the eight selected ponds based on the allowable depth of deepening, assuming a maximum depth of 3 m.

In the optimal allocation of water resources, the available water capacity of all ponds is utilized, with the remaining water being sourced from canals and groundwater. Considering that farmers may not have access to canals in certain areas, the model allows for flexibility in groundwater pumping. Therefore, in May, when there is excess water volume in the canal, part of the water requirement is met through groundwater pumping. In June and July, the optimization model allocates the available water in a way that utilizes the full capacity of canals and ponds, and the remaining water demand is supplied from groundwater sources within the allowable withdrawal limits. However, it should be noted that there is a limitation on groundwater pumping in these months to prevent the wells from drying up.

According to the optimization results, the total water shortage to supply irrigation requirements in June and July is 8.81 million and 10.43 million m³, respectively. In other words, the percentage of net irrigation requirements in these months was 84.81 and 82.22%, respectively. It is important to highlight that in May, all irrigation requirements are supplied without any limitations on water resources. Therefore, the focus of the analysis will be on the months where there is a limitation in supplying irrigation needs due to water resource constraints (June and July).

In the first scenario, the optimization model was executed by applying water allocation policies in the current conditions and increasing the depth of the ponds within the allowed boundaries. The results are presented in Table 6. The values in this table indicate that the capacity of the ponds during the rice cultivation months (May – June and July) increased from 4.09 million m³ in the current conditions to 4.42 million m³ in the first scenario. In other words, the application of the first scenario resulted in an 8% increase (0.33 million m³) in the available water volume.

Subsequently, the second scenario involved examining the first scenario with a 20% increase in efficiency through pond dredging. In the current situation, due to the growth of water-loving plants in the ponds, the available water coefficient was considered to be 50%, which reached 70% in the second scenario by increasing the efficiency by 20% due to dredging. Dredging means removing water-loving plants and other debris in the ponds. The results are shown in Table 7. The allocated water volume through ponds increased from 4.42 million m³ in the first scenario to 6.12 million m³ during the rice cultivation months. This means that the second scenario resulted in a 38.6% increase in water availability through ponds compared to the first scenario and a 49.7% increase compared to the current situation. Thus, the results indicate that pond dredging has a more significant impact compared to increasing the depth of the ponds.

The optimization model results for the first and second scenarios showed that in the first scenario, the water shortage in June and July was 8.65 million and 10.32 million m³, respectively. In the second scenario, the water shortage in these months decreased to 7.95 million and 9.72 million m³, respectively. Consequently, the percentage of irrigation requirements met in June and July was 85.08 and 82.42% in the first scenario, and 86.29 and 83.44% in the second scenario. Comparing the results to the current conditions, the first scenario led to a 0.3 and 0.2% increase in meeting irrigation requirements in June and July, while the second scenario resulted in a 1.7 and 5.1% increase in meeting irrigation requirements during these months.

In this study, we investigated the irrigation requirements of paddy fields using a water allocation optimization model that considered multiple water sources. The study area included canals, groundwater, and ponds as available water sources. Ponds were identified as a nature-based solution to address water shortages. Specifically, we examined scenarios involving increasing the volume of ponds through deepening and dredging.

The depth increase of the ponds in the scenario of deepening was determined based on the hydraulic connection limitations between the ponds and the adjacent aquifer. The results of our study highlighted the significant role of ponds as water storage tanks during drought periods. Optimization of the current conditions revealed that ponds contributed 3.4 and 3% of the required water for rice cultivation in the months of June and July when water scarcity was a concern.

The importance of ponds in drought mitigation becomes even more apparent during the rice cultivation period. The optimal results from the scenario of increasing pond volume through deepening, while considering the hydraulic connection limitations, showed a 0.33 million m³ increase in pond volume compared to the current conditions. This led to an average increase of 0.23% in water supply for paddy fields.

Furthermore, the optimal results from the scenario of increasing efficiency through pond dredging indicated a 38.6% increase in available water through ponds compared to the deepening scenario. This increase resulted in a 1.1% rise in irrigation requirements compared to the first scenario.

Therefore, in our study area, considering the hydraulic connection limitations between ponds and the adjacent aquifer, pond dredging plays a more significant role in increasing available water volume compared to pond deepening. To minimize water loss through evaporation, it is crucial to investigate the hydraulic connection between pond water levels and the groundwater levels of the adjacent aquifer when considering pond deepening.

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.