Researchers at Oregon State University have developed an irrigation management tool for planning, targeting and tracking the application of ET . It uses a comprehensive and sophisticated modeling of the disposition and fate of applied water in order to accurately project crop water availability into the future. Existing decision support systems used by growers do not incorporate energy and demand in their management strategies. Researchers at Oregon State University, Lawrence Berkeley National Laboratory, and Irrigation for the Future are collaborating on development of a decision support system that can facilitate load control automation, increased DR program participation and customer cost optimization under available electricity tariff structures. In order to do so, researchers need to develop an approach for anticipating DR event days using historical DR events, system load, and temperature data. The output from this analysis will complement the original site-specific irrigation schedule, and avoids irrigation sets from being scheduled on days with a high probability of DR. If an interruption in a planned irrigation schedule renders the original schedule infeasible, as illustrated in Example 3, the algorithm will generate alternative new schedules, reject schedules that violate operational constraints, evaluate the outcomes of feasible schedules in terms of a specified objective function,frambueso maceta and repeat this sequence in a systematic search for the best schedule. The final decision support system will provide irrigators a way to more accurately evaluate their opportunities to work with energy markets with less risk and greater transparency. This also gives energy providers and the grid a way to more accurately evaluate and predict which irrigators within their portfolios can participate in DR events as grid demands spike.
Salinity-driven reductions in agricultural production have long been a policy concern for California’s Sacramento–San Joaquin Delta . In this study we quantify the economic effects on local agriculture of changes in localized Delta water salinity for a range of sea level and water management conditions during the irrigation season. We employ the Delta Agricultural Production model , an agro-economic model for crops in the Delta that accounts for crop yield response to changes in irrigation water salinity. This work demonstrates the combined application of hydrodynamic, water salinity, and agro-economic modeling to provide policy and management insights for a major water resources problem in California. The economic effects of changes in irrigation water salinity vary in magnitude by crop, location, and the initial level of water salinity. By connecting hydrodynamic simulations with the crop production model, we find that small changes in salinity generally cause little change in Delta crop yields and revenues. Land use surveys indicate that higher-value and generally less salt-tolerant crops tend to be grown in areas of the Delta that currently have lower-irrigation water salinity; these areas do not experience major salinity changes in the simulated scenarios. These conditions allow lower-cost adaptation of cropping patterns, irrigated areas and the intensity of production factors per unit area within the Delta in response to the modeled salinity changes. Salt accumulation has affected agriculture since ancient times in Mesopotamia and Egypt, and modeling of crop salinity response has been in the literature for some decades. Crop production response to salinity also has a history in the economics modeling literature at various temporal and spatial scales . Models usually involve optimization to maximize profits or minimize costs in farming under different salinity scenarios. Also, Cardon and Letey applied Darcy’s law on “flow through a porous medium” to model plant water uptake under salinity conditions.
Knapp and Wichelns review dynamic optimization methods, finding that initial conditions matter and that large enough drainage disposal costs make water recycling more attractive. This paper uses results from Delta hydrodynamic and salinity transport modeling to provide irrigation water salinity levels for various locations in California’s Sacramento–San Joaquin Delta under a variety of sea level and water management conditions; we use these values as inputs to an agroeconomic model of crop production that includes the effects of soil salinity . Our modeling framework, presented in Figure 1, shows the flow of information among models. The hydrodynamic models provide water salinity data for different locations in the Delta. The DAP model takes crop production information from the Statewide Agricultural Production model , crop response to salinity models , and land use information from the Department of Water Resources for each water salinity scenario to produce economically optimal cropping patterns for each Delta island. Sensitivity analyses for more recent Delta export periods and fixed salinity scenarios are also part of the modeling framework. Several underlying assumptions are worth discussing. First, our approach assumes that soil salinity in the root zone is the same as that of irrigation water applied in the surface. Second, following Hoffman , we assume sufficient drainage exists in irrigated areas to avoid salt accumulation in the root zone . Hoffman concluded that many factors influencing soil salinization in general, including leaching requirements, crop salt tolerance at growth stages, shallow groundwater table, effective rainfall, irrigation efficiency and uniformity, climate, soil bypass flow, salt precipitation and dissolution, are not major factors for salt accumulation in soils in the southern Delta. In Delta locations where drainage is a concern for crop productivity, subsurface drainage has been already installed. Third, we use a sigmoidal approach for crop salinity response, as it is the best developed and well-suited for non-linear cropping optimization models like the one employed in this paper.
In addition, the sigmoidal response-function approach showed good performance compared to the threshold-linear and exponential approaches . Mass and Hoffman pioneered comprehensive assessment of crop response to soil salinity. Mass provided a threshold approach in which different crop types have relative yields constant up to thresholds in soil salinity. Beyond a threshold, relative yields decline at a constant rate. Another approach describes crop response to soil salinity in the root zone using a sigmoidal function that calibrates to a soil salinity at which crop yields are reduced by 50 percent. Other factors that may affect crops include drainage and irrigation water salinity. Drainage salinity is closely related to soil salinization, because poor soil drainage conditions retain salts. A rising groundwater table with brackish or saline water can degrade soil at the root zone with prolonged exposures. Salinity in irrigation water decreases yields for many crops. However, brackish or slightly saline irrigation water may not affect yields for some crops if the appropriate drainage exists, in which case salts do not accumulate in the root zone. Below, we present the DAP model structure and data sets, the water salinity scenarios and hydrodynamic modeling work, and model results for the water salinity scenarios . We conclude with a summary of the findings and some suggestions for further research.The Delta Agricultural Production Model estimates the irrigated crop area and the crop mix that maximizes total net revenues on land areas within the Delta, taking into account production costs, crop prices, water use,cultivar frambuesas and yield changes from irrigation water salinity . DAP is a customized version of the SWAP agro-economic model of California agriculture, augmented to examine the effects of irrigation water salinity. SWAP uses positive mathematical programming to calibrate a base case to observed values of input use, namely land, water, labor and supplies. SWAP has been used for numerous agricultural modeling applications in California including water markets, soil salinity in the Central Valley , climate change , and regional economic impacts of water markets and drought in the Central Valley . The hydrodynamic modeling used to estimate salinity changes of Delta waters is based on two models developed by Resource Management Associates, Inc. for the state-commissioned Delta Risk Management Strategies study and reported in Fleenor et al. . Development, verification, calibration and validation of both models can be found in Fleenor et al. , Bombardelli et al. and , and Fleenor and Bombardelli . First, the one-dimensional Water Analysis Module is used to estimate salinity changes with the introduction of dual conveyance of water exports and sea level rise. Fleenor et al. performed simulations with WAM over water years 1981 through 2000. Second, a two-dimensional RMA Bay-Delta model was used to estimate salinity changes from the permanent flooding of five western islands that serve as a salinity barrier at the Delta’s western edge .
RMA performed these 2-D simulations that spanned the April 2002 through December 2004 hydrologic period for the DRMS study. Permanent flooding represents conditions where the islands have either been flooded for some time or during winter months when considerable freshwater flows are available, but not the near-term results of a “Big Gulp” of salt water flowing into the Delta that might occur with catastrophic island failures during the summer or fall. We summarize these modeling results and show the model output water salinity sampling locations . To assign irrigation water salinity for each island and water salinity scenario we located the two closest sampling locations and then selected the sampling station with the highest monthly average salinity during the irrigation season. The supplementary tables in the project website provide detailed information on the sampling stations used and simulated monthly average salinity levels by island and hydrodynamic modeling scenario. To account for the largest possible monthly average salinity levels, we explored salinity conditions within a relatively long irrigation season . This choice also likely overstates the average salinity conditions most farmers face when irrigating their crops, because salinity tends to be highest in the late summer and fall, when most irrigation is finished except for pasture and hay crops.WAM simulations contrast 1981–2000 salinity conditions for three sea levels . The sea level rise projections are within the range the California Ocean Protection Council recommends for long-term planning purposes, based on recent model projections for the mid- and late-21st century . Some projections anticipate the potential for higher sea level rise by the end of the century, and these would likely generate higher salinity levels than those shown here. WAM simulations also include two Delta export configurations . RMA 2-D simulations contrast a 2002–2004 base salinity case with all islands intact and a scenario with five western islands flooded , the hatched area in Figure 3. RMA 2-D runs do not consider sea level rise. For WAM, we also contrast a base case and a dual conveyance case for critically dry years within the modeled time period . For both WAM and RMA 2-D runs, all cases assume the same daily hydrology and water system operations as those which actually occurred during the modeled periods. In the case of dual conveyance, the model draws exports through the new conveyance system unless these exports would cause Sacramento River flows to fall below a minimum environmental flow of 283.2 m3 s-1 . This environmental constraint is introduced to avoid reverse flows at the intake points that could harm fish . Average export levels during the 1981–2000 reference period used for WAM were 5.96 billion m3 yr-1 , and 7.14 million ac-ft yr-1 for the 2002–2004 reference period used for RMA 2-D. Reference salinity for each hydrodynamic model run are shown in supplementary tables in the project website . Figure 4 shows salinity as electrical conductivity during the irrigation season for the five agricultural sub regions within the Delta under different export conveyance and sea level rise cases.At current sea level, dual conveyance would increase salinity in most regions , though not necessarily in the eastern and central parts of the Delta . Sea level rise increases salinity in most cases . However, dual conveyance operations combined with sea level rise may not increase salinity in the eastern and central Delta . During dry years, salinity is generally higher than during other years in the modeled time period, and dual conveyance increases average salinity at least marginally in all regions in both the irrigation and non-irrigation seasons, as shown in Figure 5. The permanent flooding of western islands does not result in large increases in salinity over the base case during the irrigation season, although it does increase salinity somewhat more in the non-irrigation season .