Irrigated agriculture is crucial for global food production, particularly in snow-dependent regions. This study focuses on the Yakima River Basin (YRB), a major agricultural area in the Pacific Northwest of the US, heavily reliant on snowmelt for irrigation. Climate change impacts this system through two primary mechanisms: direct biophysical effects on crop growth (altered CO2 concentrations, temperature, and precipitation) and indirect effects through changes in water availability due to altered snowpack dynamics and increased drought frequency and severity. The research questions are twofold: 1) What are the direct and indirect impacts of climate change on irrigated agriculture in the YRB? 2) To what extent does farmer adaptation (through improved crop varieties) mitigate the negative effects of climate change? The study considers both the expected mean yields and interannual yield variability, recognizing the significant implications for crop insurance programs and revenue stability.

The introduction cites existing literature demonstrating the importance of irrigation in global food production and the dependence of many regions on snowmelt for water supply. It highlights the biophysical effects of climate change on crop yields, including the positive effects of increased CO2 and the negative impacts of warming temperatures on crop growth and productivity. The literature also shows the expected increase in drought severity and frequency in snow-dependent areas due to reduced snowpack and altered snowmelt patterns. The paper points out limitations of previous simulation-based assessments that don't fully account for regulated flows, institutional water rights, and regional regulations.

The study employs a physically-based, coupled agro-hydrologic platform (VIC-CropSyst and YAK-RW) to simulate agricultural and hydrological processes within the YRB. VIC-CropSyst models crop growth, development, and water interactions, while YAK-RW simulates dam operations and water allocation under various scenarios. The model operates at a 1/16-degree spatial resolution, covering the YRB, with a historical period from 1980-2010 and future projections from 2030-2090. Climate inputs are derived from five general circulation models (GCMs) under two representative concentration pathways (RCPs) 4.5 and 8.5. Soil, land cover, and crop parameters are derived from existing datasets. The study defines an "unmet demand" (UD) index to represent water stress, which is calculated as (1 - Water supply/Water demand). The model simulates the yield of ten major crops in the YRB, categorized into annual crops, multiple-cutting crops, and tree fruits. The study considers two scenarios: a status quo scenario using current crop varieties and an adaptation scenario incorporating improved crop varieties with longer growing periods. The model is used to evaluate the average annual yield, interannual yield variability (standard deviation), and the probability of low-yield years under different climate and adaptation scenarios. The YRB's unique water rights system, which includes proratable and non-proratable water rights, is incorporated into the model.

The simulations show a significant increase in water stress (UD) in the YRB under future climate scenarios, primarily driven by temperature changes. The probability of significant droughts increases substantially. All crop types show vulnerability to climate change, with significant potential reductions in average annual yields. For example, by 2060-2090 (RCP 8.5), average annual yields of potato, alfalfa, and apple could decline by 46%, 22%, and 48%, respectively. While some multiple-cutting crops may benefit from higher temperatures and increased growing cycles, this is outweighed by the impacts of water scarcity. Interestingly, the interannual variability of yield does not necessarily increase with more frequent droughts because maximum yield decreases for many crops, narrowing the yield gap between fully irrigated and non-irrigated conditions. This has implications for crop insurance, potentially leading to lower premiums with current varieties but higher premiums with improved varieties. The adaptation scenario using improved crop varieties shows increased average yields but significantly higher interannual yield variability. The probability of low-yield years substantially increases under both RCPs. This creates a mean-variability tradeoff: improved varieties increase average yield but also increase risk and revenue volatility. This increased volatility has implications for insurance programs and government subsidies.

The findings highlight the complex interplay between climate change, water availability, and agricultural productivity in snow-dependent regions. The study demonstrates that focusing solely on biophysical adaptations (e.g., improved crop varieties) may not be sufficient to ensure sustainable agricultural productivity. Addressing water availability through improvements in infrastructure and water management institutions is equally crucial. The mean-variability tradeoff underscores the importance of a holistic approach considering both biophysical and institutional factors when developing adaptation strategies. The implications for crop insurance are significant, highlighting the need for adjustments in premium rates and government subsidy programs given the changing yield variability.

This study underscores the need for simultaneous strategies to enhance both water availability and crop productivity in snow-dependent regions. Improvements in water infrastructure and institutions are critical, coupled with careful consideration of the tradeoffs associated with crop adaptation strategies. Future research should further explore diverse seed varieties and faster breeding processes to enhance climate resilience. A more refined simulation of crop heat stress and its dependence on irrigation water for both cooling and crop water demand is also warranted.

The study does not consider factors like changes in cropping patterns, geographical shifts in agricultural areas, changes in irrigation systems, or potential future changes in water rights regulations. The model does not capture crop mortality due to heat stress. The adaptation scenario only explores one type of improved crop variety (slower-maturing varieties), and other strategies might yield different results. While the general conclusions are broadly applicable, the specific quantitative projections should be considered within the limitations of the modeling framework.