SEASONAL CLIMATE FORECASTS: SKILL AND UTILITY FOR DECISION MAKING IN AGRICULTURE AND WATER RESOURCES MANAGEMENT.

Ashish Kondal

doi:10.7273/000007234

Seasonal climate forecasts, looking out the next several months, are a potential solution to lessen the adverse impacts and uncertainty associated with everchanging climate conditions in agriculture and water resource management sectors. Despite their availability, these forecasts are still underutilized which is attributed to a lack of availability of forecasts of decision-relevant variables at appropriate decision timeframes for desired locations and a perception of poor skill by decision-makers. This dissertation focuses on tackling the aforementioned impediments to the utilization of seasonal climate forecasts in the agriculture and water resource management sectors. Our research first investigated the seasonal forecast skill for agriculturally relevant variables derived from a multi-model ensemble forecasting system and subsequently demonstrated the skill of these forecasts in informing key agricultural decisions. Additionally, this dissertation quantified changes to water market outcomes in drought contexts, when agents participating in water markets make decisions under imperfect drought forecasts rather than perfect forecasts. In the first essay, we performed a comprehensive skill assessment of agriculturally relevant variables ranging from the meteorological drivers (temperature and precipitation), water supply, agricultural water demands, and crop performance, within our case study region i.e., the Pacific Northwest Region of the United States. We observed positive skill in forecasting these variables, with higher skill in predicting the upper tercile category regardless of the variables assessed. We observed spatiotemporal variability and variability across crop groups in skill, highlighting the space-time dependency of forecast accuracy. This allowed us to identify high-skill regions and timeframes, which form the foundation for real-world applications. Building on the skill identified in cumulative runoff and temperature forecasts in the first essay, we demonstrated its application in informing fall fertilization and drought response decisions in our second essay. We found a positive forecast skill at least two months in advance of when the decisions are typically made, with higher accuracy for locations that are critical to the respective decisions. Our third essay focused on analyzing the impact of forecast error on estimates of aggregated economic outcomes resulting from participation in water markets. Using a baseline agent-based model, we quantified the relative change in economic gains derived from agents having to base decisions on imperfect forecasts of drought rather than perfect forecasts. Through this dissertation, we contributed to the advancement and applicability of seasonal climate forecasts in agriculture and water management sectors. We highlighted the spatiotemporal variability in forecast skill, demonstrating the non-uniformly distributed skill across space, time, variables, and crop groups. This finding is pivotal for tailoring seasonal forecasts to specific regions, decisions, and timeframes, thereby improving the reliability and relevance of agricultural decision-support systems. From a decision-making perspective, we bridge the gap between theoretical forecast skill and practical application. By leveraging the identified skill in cumulative runoff and temperature forecasts, our study illustrates how these forecasts can be effectively used to inform critical agricultural decisions, such as fall fertilization and drought response. This showcases the direct applicability of the research findings in real-world scenarios, allowing decision-makers to be more proactive and better prepared to make informed actions based on provided seasonal forecasts for the locations critical to specific agricultural decisions. Furthermore, we investigated how do water market outcomes vary with the integration of seasonal climate information into an agent-based model for water market decisions. This analysis provides valuable insights into how forecast errors affects the economic outcomes of a decision, offering a nuanced understanding of the risks and benefits associated with the use of imperfect forecasts in water resource management. Collectively, our dissertation lay the groundwork for future research and the development of more refined seasonal forecasts enabled decision-support tools in agriculture and water management sectors.

SEASONAL CLIMATE FORECASTS: SKILL AND UTILITY FOR DECISION MAKING IN AGRICULTURE AND WATER RESOURCES MANAGEMENT.

Files and links (1)

Abstract

Metrics

Details