Date of Award

3-2022

Document Type

Thesis

Degree Name

Master of Science

Department

Department of Systems Engineering and Management

First Advisor

Justin D. Delorit, PhD

Abstract

Natural hazards, such as hurricanes, wildfires, floods, and droughts impact human systems that rely on predictable patterns in the natural elements with which they interact. Skillful prediction of the impacts of climate change on linked, human-natural systems, like surface water resources, can help ensure physical risks within vulnerable communities are mitigated, resource sustainability is maximized, and intersectoral markets continue to contribute to socioeconomic stability. Due to water resources being a primary conduit through which climate uncertainty impacts people, economies, and ecosystems, its study is worthy of investigation; particularly, where those resources are uncertain and demanded by a variety of competitive users. This work evaluates a season-ahead statistical prediction model of growing season streamflow (September – December) in Andes, Antioquia, Colombia, against a suite of global and local predictor variables: precipitation, soil moisture, Niño 3.4 sea-surface temperature anomaly, and Southern Oscillation Index anomaly. Skillful results, which are defined as streamflow forecasts that outperform a specified climatological baseline, are produced for the models when analyzing extreme streamflow events (r^2 = 0.77, mean absolute percentage error = 21.87, ranked probability skill score = 0.21). Viewed cumulatively, these results suggest streamflow predictions and forecasts can identify the role of global and local climate on communities, inform how and when changes should be implemented, and justify stakeholder decisions.

AFIT Designator

AFIT-ENV-MS-22-M-252

DTIC Accession Number

AD1174722

Share

COinS