The Pajaro Valley Water Management Agency formed in 1984 to manage existing and supplemental water supplies within the basin. PVWMA has implemented policies and an assortment of strategies to address groundwater overdraft while maintaining agricultural productivity and meeting water demands in the area, which have risen steadily in the past 50 years along with population, agricultural acreage, and groundwater extraction, with pump age rising from 7.5 million m3 in 1964 to 13.5 million m3 in 2009 . Agricultural land predominates in the valley and was estimated at 10,000 ha in 2006, compared to 5,000 ha for urban and rural municipalities . Crops include berries, vegetable row crops, grapes, apples, and cut flowers, and production has developed into a multi-million dollar agricultural sector with crop yields valued at over $800 million in 2011, and the region ranks fifth for total agricultural production in California . Large corporations, such as Driscoll’s, California Giant, and Martinelli’s & Company helped the area to become one of the top ranked farming cities in the country, and this agroindustrial pressure creates a unique economic environment worth studying, which includes high-value crops, a recycled water system, and an aquifer recharge basin.Although crops shift yearly due to numerous factors such as traditions, preference, economic profit, etc., the total area of agricultural land has remained consistent since 1989 . Agricultural demand for Pajaro Valley was divided into inland and coastal regions. Given the extent of the historic seawater intrusion and coastal access, agricultural demand for Pajaro Valley was divided into inland and coastal regions that are delineated by the CDS and Highway 1 . This study defined inland crops as strawberries, vegetables , bush berries , vine grapes, artichokes, apple trees, cut flowers, and other crops and coastal crops as strawberries,ebb and flow trays vegetables, artichokes, cut flowers, and a small number of other crops. These crop assignments were adapted from the 2012 PVWMA data in the Basin Management Plan Update .
Overdraft of the Pajaro Valley groundwater basin has depleted the aquifer storage and led to saltwater intrusion from Monterey Bay into freshwater aquifers, causing water quality degradation and unsustainable storage levels. Seawater intrusion has been observed up to 4.8 km inland and could potentially reach farther if over extractions continue . PVWMA is executing several measures as part of a Basin Management Plan to address the imbalance of water demands and supplies. The CDS was implemented in 2009 to supply irrigation water to farms in coastal areas with compromised groundwater supplies. Water from the CDS serves in lieu of local groundwater and helps to reduce coastal seawater intrusion by reducing groundwater pumping near the coast through the delivery of a mixture of groundwater from farther inland in the basin, water recovered from a locally managed recharge system, and wastewater from the Watsonville Area Water Recycling Project. This facility and its conveyance system can produce ∼4.934 million m3=year , which includes recycled water, Harkin Slough recovery wells, and blend wells. PVWMA also increased groundwater supplies through the Harkins Slough Project, a managed aquifer recharge and recovery basin. The project aims to replenish a shallow aquifer by infiltrating water diverted from the Harkins Slough in the winter to provide an alternative solution to the overuse of groundwater. Water demand and supply challenges of this area provide a unique opportunity to develop strategies for improved water allocation and conservation. This study uses the coupling of a simulation and optimization model to provide a unique approach to sustainable groundwater management and could be integrated into future decision-making processes and groundwater sustainability plans.A GBM was built and calibrated to represent the water supply, water use, and groundwater storage of Pajaro Valley groundwater basin. An optimization model was built to determine crop acreages that maximize agricultural profit given water availability constraints. The models were coupled to estimate the aquifer storage and assess the sustainable carrying capacity of Pajaro Valley groundwater basin by using the outputs from the optimization model as inputs into the groundwater box model.Data sources for inflows included land use data and Harkin Slough recharge inflows obtained from PVWMA and precipitation and reference evapotranspiration obtained from the California Irrigation Management Information System. An estimate of evapotranspiration was developed based on data from the National Solar Radiation Database and the Hargreaves-Samani equation whenever ETo data was unavailable . Monthly crop coefficients values were obtained from Hanson et al. . In this study, the agriculture water demand was estimated using potential evapotranspiration, which can differ from the actual evapotranspiration. The application efficiency was assumed to be spatially uniform throughout the valley.
The percentage of irrigation use for gravity, sprinkler, drip, and other methods were obtained from the California Department of Water Resources and average application efficiencies were obtained from Sandoval-Solis et al. . Data sources for outflows include population data for the City of Watsonville, and rural municipalities that were retrieved from the US Census Bureau , and water use per capita for indoor and outdoor consumer water use from 1999 to 2015 was obtained from Cahill et al. . WUPC from 1966 to 1999, was assumed as the fixed value of the 1999 WUPC. Similarly, WUPC from 2016 to 2040 was assumed to be the same value as in the year 2015. Rural WUPC was estimated to be 29% of the City of Watsonville, based on the urban-rural population ratio. The acreage factors and acreage share percentage were obtained from Lin et al. . Well production data for the City of Watsonville, agricultural wells and recycled water was provided by PVWMA. Linear regression models were used to fill gaps when input data for specific periods were missing. Table S1 shows the model equations.Based on Eq. , at a given time, a groundwater basin has a certain amount of water that might increase or decrease based on the change of storage. If the total inflows are greater than the total outflows, the positive change will increase the groundwater storage. Conversely, if the outflows exceed the inflows, then the negative change will result in the decrease in groundwater storage. In this study, net groundwater storage is defined as the average change of storage for a determined period, and the change of storage is calculated every year by subtracting the inflows minus the outflows. GBM inflows ranged from 24.6 to 96.6 million m3=year, and outflows ranged from 40.7 to 98.6 million m3=year. In contrast, PVHM inflows ranged from 16 to 103 million m3=year, and outflows ranged from 30.8 to 90 million m3=year. Groundwater pumpage is dominated by agricultural use and was 13.5 times greater than urban and rural water demands. Recharge to the aquifer from precipitation is 6.2 times greater than recharge due to excess irrigation. Fig. 2 shows the net change in groundwater storage of the GBM and PVHM.The optimization model was built to estimate a series of optimal acreages that maximize economic profits. Profits were estimated as yearly benefits , which were the difference between crop revenue and the costs of production for that crop. All prices were adjusted using the consumer price index for 2015. Crop revenues were calculated based on crop incomes from crop reports and economic contributions of Monterey County from 1966 to 2014 and the annual crop and livestock reports of Santa Cruz County base.
COP components of the model were obtained from budgets published through current cost and return studies from the University of California Cooperative Extension . These budgets were used to determine annual costs per acre for each crop including operational, cultural, and overhead costs that covered land preparation, plant establishment, fertilization, pest management, harvest, labor, equipment costs, property taxes, irrigation, sanitation, and management salaries. The price of water was removed from each budget because it was included separately in the optimization equation for specific circumstances in Pajaro Valley. COP budgets for crops grown in the California central coast region were used for strawberries, vegetables, bush berries, artichokes, and apple trees, while the COP budget for grapes grown in the upper San Joaquin Valley was used for vine grapes. The UCCE has not published a COP budget for cut flowers in California; therefore estimates were made based on COP budgets developed by the Cooperative Extension at Penn State University. Based on budget availability, each crop was treated slightly differently. The COP for strawberries was found by taking the average of two budgets, one for each year of production . The COP for vegetables was determined based on budgets for various types of lettuce grown in this region because lettuce is commonly used in rotation with strawberries . The COP for bush berries was calculated through consideration of the budgets for raspberries and blackberries,4×8 flood tray including an establishment year, the first year of reduced production, and four subsequent years of steady production . The COP for artichokes was based on a single production year . The COPs for apples and vine grapes were based on a 25-year life of an apple orchard and vineyard, respectively . The first year of establishment for vineyards is the most expensive, with subsequent years costing one-third of the initial price. The COP of the other crop group was based on the budgets of alfalfa, wheat, and beans . Table 2 shows the COP budgets and revenue figures. All crops show economic benefit except for vine grapes, for which revenue remains below the break-even point. From the year 2000 to 2010 vine grapes had economic benefits, but not from 2010 to 2015 where COP exceeded revenues. Vine grapes have become less economically viable in recent years because of inexpensive imports from Australia, competition from corporate farms in other regions of California, and the fact that harvest standards often change in harvest time, hang time, and Brix standards .
The cost of water in Pajaro Valley was defined by the price of water and the energetic cost for pumping. Annual rates published by the City of Watsonville and PVWMA set water prices, which differ based on user location and water source . Water prices ranged from $101 to $338 per acre-ft and increased with an average yearly rate from 3% to 9.5% . The energetic cost of pumping was estimated to range from $0.18 to $0.20 kWh for an average well depth of ∼90 m for domestic wells, and ∼131 m for municipal wells .The coupled model determined the optimal crop pattern by maximizing net economic benefits while constraining agricultural water and land use, which decreased groundwater overdraft. Both scenarios began in the year 2000 with approximately ∼8,000 ha and 62.5 million m3=year of water use. For the next 15 years, the trend increased to ∼8,500 ha and 64.1 million m3=year for the baseline scenario and these values were maintained until 2040. The optimized scenario decreased water use to 49.3 million m3=year for ∼6,315 ha . The crop acreages that gained the most economic revenue and water use were bush berries, cut flowers, strawberries, and vegetables. These crops were allocated within their maximum allowable acres. The lowest crop revenues were for apple trees, vine grapes, artichokes, and others, reflecting their minimized acreages. The optimal land use had a total acreage reduction of 15%. The objective function was to maximize the net revenue from agricultural production while determining the optimal crop pattern for the available water. A similar objective function is observed in other studies. Mainuddin et al. determined the irrigation plan by optimal crop area allocation and groundwater requirement by maximizing the net economic benefit in Thailand. Benli and Kodal developed a linear model that allocates optimally available resources, rearranges crop patterns, and maximize economic crop revenue. These studies showed a decrease in available water corresponded to the upper limits of acreages of higher values crops, which is consistent with the finding from this study that crop acreages increase or decrease relative to the change in their economic profit. Regarding economic benefit, the baseline scenario showed total revenue of $274 million, which increased to $289 million in the optimized scenario using the same allotment of water . The net difference between the optimized and baseline scenario is shown in Table 4, where the pattern of higher revenue for the optimized scenario is observed even when the available water was reduced to 60 million m3. However, if available water is reduced to 50 million m3, revenue is reduced to ∼$239 million. In general, annual revenue decreased by 2.4% on average, which translates to ∼$5 million loss per reduction of 1.2 million m3 of available water.