Back to search

Cornell University

SUSTAINABLE NUTRIENT MANAGEMENT THROUGH MODELING AND FORECASTING FRAMEWORKS FOR WATER QUALITY DECISION SUPPORT

Abstract

dc:description.abstract

Nutrient pollution from agricultural and urban sources continues to degrade freshwater ecosystems worldwide. Phosphorus and nitrogen, particularly in their bioavailable forms, drive these processes, with phosphorus often acting as the primary limiting nutrient in freshwater systems. While best management practices (BMPs) such as reduced fertilizer use, cover crops, and riparian buffers aim to limit nutrient losses, their effectiveness is highly variable and increasingly challenged by climate change, extreme precipitation, and shifting hydrology.This dissertation investigates nutrient management at multiple scales in New York State, focusing on both watershed modeling for long-term planning and short-term, forecast-informed decision-making. Chapter 1 presents an alternative SWAT model calibration approach for a data-limited watershed (Canandaigua Lake, central New York) and evaluates the effects of calibration uncertainty on estimated BMP effectiveness. Results show that total phosphorus (TP) loads, their spatial distribution, and both absolute and percent load changes vary under parameter uncertainty; however, the relative ranking of BMP effectiveness remains largely consistent. Chapter 2 builds on this by using a model-as-truth experiment in the Cayuga Lake watershed to evaluate how differences in data availability influence model performance and BMP assessment. The findings demonstrate that TP load estimates and estimated changes in load under BMPs vary considerably across data availability scenarios, but the amount of data available for calibration does not explain how close model estimates approach their true values. This outcome is largely attributed to the challenge of calibrating a high-dimensional model in the presence of significant equifinality. Still, the ranking of BMP effectiveness remains stable across data availability scenarios, consistent with Chapter 1 results. Chapter 3 evaluates the skill of short-term probabilistic runoff forecasts across New York State, a tool meant to inform short-term, farm-scale decisions around manure and fertilizer application. Results show that the runoff forecasting system effectively identifies snowmelt- and thaw-driven runoff events but tends to underestimate low-probability and rainfall-driven events, with forecast performance varying across watersheds. Overall, this work demonstrates that combining robust calibration, uncertainty quantification, and real-time runoff forecasting can support adaptive nutrient management. Future research should explore alternative calibration methods and machine-learning post-processing to improve model reliability and operational decision support

Degree

thesis:*
Name thesis:degree_name
Ph. D., Biological and Environmental Engineering
Level thesis:degree_level
Doctor of Philosophy
Discipline thesis:degree_discipline
Biological and Environmental Engineering
Grantor
Cornell University
Year dc:date.issued
2025

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Sepehrmanesh, Mahnaz
Chair dc:contributor.chair
  • Steinschneider, Scott
Committee members dc:contributor.committeemember
  • Jung, Sunghwan
  • Anderson, Catherine

Subjects

dc:subject × 6

Rights

dc:rights
Statement dc:rights
  • Attribution 4.0 International
Language dc:language.iso
en

Identifiers

dc:identifier.*
Dc Identifier Bibid
17423280
Dc Identifier Other
ProQuest Submission ID: 15348
ProQuest Publication ID: 32285602

Chain of custody

source
Harvested from
Cornell University
Base URL
ecommons.cornell.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Sepehrmanesh, Mahnaz. SUSTAINABLE NUTRIENT MANAGEMENT THROUGH MODELING AND FORECASTING FRAMEWORKS FOR WATER QUALITY DECISION SUPPORT. Doctor of Philosophy thesis, Cornell University, 2025. https://hdl.handle.net/1813/121069