Recipient Organization
VIRGINIA POLYTECHNIC INSTITUTE
(N/A)
BLACKSBURG,VA 24061
Performing Department
Biological Systems Engineering
Non Technical Summary
Climate variability already strongly influences agricultural yields, and climate-related disruptions to agricultural production have increased in the last 40 years. For instance, yields for crops such as soybeans and peanuts in the 2007 drought year in Virginia were approximately half of normal levels. Heat waves can also lead to production shortages that cannot be addressed through irrigation alone. For example, one farmer reported in 2010 that he "can't water his way out of the heat". Furthermore, many Mid-Atlantic States that are accustomed to ample water resources are experiencing increasing competition for water from population and industrial growth, meaning that unlimited water for irrigation may not always be available. Climate change has the potential to exacerbate these challenges, leading to negative consequences if adaptive measures are not taken. For instance, one study found that warmer temperatures over the past three decades have resulted in a 5.5% reduction in global wheat production relative to a counterfactual without higher temperatures, with reductions of up to 15% in some regions of the world. Some research suggests that non-irrigated humid areas of the US are particularly vulnerable relative to irrigated arid regions.Effective management of these risks requires a firm understanding of the relationship between climate and agricultural production and the ability of producers to adapt and respond to climate risks. This requires regional-scale analyses, combined with data on farming and irrigation practices, to appropriately characterize vulnerabilities and risks for agricultural systems in the Southeast. It is important to recognize that climatic risks and risk management measures will interact across multiple scales in a manner that field-scale research alone cannot address. For instance, the expansion of irrigation could reduce short-term risks to individual farms but lead to a long-term decline in water availability if irrigation exceeds water availability. Understanding and managing these interactions will be important to policy makers and producers aiming to reduce the impacts of climate change.This project will integrate research on the interactions between climate, agricultural production, farmer decision making, and water supply policies to provide a comprehensive picture of climate-related risks to agriculture in humid regions. This will improve our understanding of climatic risks to agriculture over both the near and long-term, and will also provide insights into the degree to which adaptive measures at the farm, state, and regional-level can address these risks. Ultimately, this work aims to reduce the impact that climate variability has on agricultural production. This work aims to improve the capacity for producers and policy makers to assess and manage long-term risks associated with climate change. While the research will be focused on Virginia specifically, the results and methodologies developed will be applicable to the eastern United States.
Animal Health Component
40%
Research Effort Categories
Basic
60%
Applied
40%
Developmental
(N/A)
Goals / Objectives
The overall goal of this work is to improve our understanding of climate-related risks to agriculture in humid regions and evaluate the effectiveness of different risk mitigation measures. This research will consist of four phases aimed at achieving the following specific objectives:1. Statistical climate-crop yield modeling: Characterize the statistical relationship between agricultural yields in Virginia and various climate parameters using fixed-effects panel regression and automated variable selection.2. Agricultural water use evaluation: Evaluate typical irrigation management methods for Virginia producers and the influence that climate conditions have on irrigation management methods and water usage.3. Integrated modeling of climate, hydrology, water use and agriculture: Evaluate how producer irrigation practices interact with changes in water availability to influence watershed-level risks associated with various climate change scenarios.4. Robust decision making assessment: Assess the degree to which robust decision frameworks can inform risk mitigation policies and incentives aimed at reducing climatic risks to agriculture. Ultimately, I aim to use the data and analyses developed for Virginia as a foundation for pursuing additional external funding (for instance, from USDA or NSF) that evaluates the relationship between water, climate and agriculture across humid regions more broadly.
Project Methods
Objective 1: Statistical Climate-Crop Yield ModelingThis will be done through an empirical evaluation of the statistical relationship between historic climate data and agricultural yields for six crops (corn, cotton, hay, tobacco, soybean, and wheat).Data on crop yields will be taken from the USDA National Agricultural Statistics Service (NASS), which collects annual yield data from all counties in the United States.Historic daily climate time series for each county will be extracted from the PRISM climate dataset,which is available at a 4-km spatial resolution making it suitable for county-level analyses. PRISM includes daily time series from 1981 through the present for temperature and precipitation. Daily data (particularly on precipitation) is necessary to evaluate the relationship between climatic variability and crop yields, which can't be assessed based on monthly- or annually-aggregated data. This will allow inclusion of more nuanced climatic explanatory variables, such as consecutive days above 30°C, consecutive days without rainfall, and rainfall intensity. To account for variations in average climate condition in different geographic regions of the state, all climate variables will be converted to anomaly values that represent how temperature and precipitation in a given year compared to the long-term average for that county.The relationship between climate conditions and detrended yields for each of the six crops will be assessed via a fixed effects panel regression, whichaccounts for variation in unobserved attributes of the different counties (such as soil type, topography, or socioeconomic conditions), and thus is not subject to omitted variable bias that can limit interpretation of cross-sectional studies.Objective 2: Agricultural Water Use EvaluationThis will be accomplished first by surveying producers on irrigation practices, and second through an empirical assessment of historic water use data that has been reported to the Virginia Department of Environmental Quality (VDEQ).In the first part, a written survey on farm-level irrigation practices will be developed to collect data on typical irrigation management practices for different crops. This survey will collect information on:Basic information about farm operations, such as acres farmed, crops grown, land ownership status, etc.Water sources and irrigation equipment utilized, including diversions.Methods used in deciding when to irrigate, including specifics on scheduling approach and the specific crop and soil conditions that prompt water application.Degree to which climate conditions and forecasts impact irrigation scheduling decisionsBarriers to the adoption of methods for conserving irrigation water and energy.The survey will be developed in coordination with specialists and agents within Virginia Cooperative Extension (VCE) to ensure that the questions are clear and understandable. To encourage a high response rate, the survey will be administered in coordination with in-person VCE workshops as well as commodity-group meetings.This phase of research will also evaluate how reported agricultural water use varies with climate conditions by developing statistical models that relate historic agricultural water use with climate variables. This will closely mirror the analysis conducted for Objective 1 of the project, but the response variable in this case will be irrigation water use data that has been reported to VDEQ.Objective 3: Integrated modeling of climate, hydrology, water use and agricultureObjective 3 will develop an integrated systems dynamics model that simulates climate risks to both water supply and water demand, as well as farmer irrigation practices identified in Objective 2 of this project. This integrated model will be used to evaluate how surface water availability responds to different scenarios of climate change and irrigation development. Results from this model will be compared to those from a simplified model that assumes irrigation is applied scientifically (in response to reference crop evapotranspiration) to evaluate the degree to which irrigation practices influence long-term risks to water availability and crop production.Physical hydrology will be modeled using the Soil and Water Assessment Tool Variable Source Area (SWAT-VSA)39 model for the South Fork of the Shenandoah River watershed. This model simulates runoff from hydrologic response units (HRUs) based on climatic, soil, land-use, and topographic conditions, and incorporates a topographic wetness index to account for variable source areas (VSAs) that are the predominate contributors to surface water runoff.Two components will be added to the model to represent the interactions between surface water flow, irrigation, climate conditions, and agricultural production. The first component will model irrigation water withdrawals from surface water that dynamically respond to climate conditions based on the findings from Objective 2 of this project. The second additional component will incorporate crop yield response models developed for Objective 1 of the research project. County-level crop yields will be estimated as a function of climate conditions and agricultural crop cover from the Cropland data layer.Objective 4: Robust Decision Making AssessmentIn this phase of the project, Robust Decision Making (RDM) will be used to assess various options for mitigating climate-related risks to agriculture identified for Objective 3 of the project. RDM uses simulation models to assess how a system performs in many possible future states of the world. The results of these simulations are stored in a database, which is then analyzed to identify which alternative performs satisfactorily in the largest range of potential conditions, and the specific conditions that result in unsatisfactory performance for each alternative.The first step of the RDM analysis requires identifying alternative policies aimed at reducing the risk of water shortages and drops in agricultural production. Specific policies will be determined after potential impacts are identified for Objective 3, but possibilities include reduced allocation of withdraw permits, restrictions on water withdraws under specific low-flow conditions, and incentives for improved efficiency. Following this, we will identify the uncertain conditions that could impact water supply and agricultural productivity in the future. After these uncertain parameters have been identified, the model developed for Objective 3 will be used to simulate water supply and agricultural production over many randomly sampled combinations of uncertain parameter values. In each simulation, each policy alternative will be compared based on agricultural production and the frequency of surface water flows below minimum thresholds identified in the Virginia State Water Resources Plan.The simulation results will be assessed to identify which policy alternatives offer the most robust performance with regard to agricultural production and maintaining minimum water flows. For each alternative, the results will also be reviewed to identify the uncertain parameters that have the strongest influence on agricultural productivity and water flows, through a process known as "scenario discovery." This part of the analysis identifies the uncertain parameters that are most consequential for overall performance, which provides a characterization of specific vulnerabilities and insights into places where a reduction in uncertainty would be most valuable.