Evaluation of uncertainty in a Maumee River Watershed Soil and Water Assessment Tool under current conditions and future climate projections

Evaluation of uncertainty in a Maumee River Watershed Soil and Water Assessment Tool under current conditions and future climate projections

Bibliographic Details
Main Author: Kujawa, Haley A.
Language:English
Published: The Ohio State University / OhioLINK 2019
Subjects:
Environmental Science
SWAT
climate change
uncertainty
watershed modeling
model validation
Online Access:http://rave.ohiolink.edu/etdc/view?acc_num=osu1555575109524802
id ndltd-OhioLink-oai-etd.ohiolink.edu-osu1555575109524802
record_format oai_dc
collection NDLTD
language English
sources NDLTD
topic Environmental Science
SWAT
climate change
uncertainty
watershed modeling
model validation
spellingShingle Environmental Science
SWAT
climate change
uncertainty
watershed modeling
model validation
Kujawa, Haley A.
Evaluation of uncertainty in a Maumee River Watershed Soil and Water Assessment Tool under current conditions and future climate projections
author Kujawa, Haley A.
author_facet Kujawa, Haley A.
author_sort Kujawa, Haley A.
title Evaluation of uncertainty in a Maumee River Watershed Soil and Water Assessment Tool under current conditions and future climate projections
title_short Evaluation of uncertainty in a Maumee River Watershed Soil and Water Assessment Tool under current conditions and future climate projections
title_full Evaluation of uncertainty in a Maumee River Watershed Soil and Water Assessment Tool under current conditions and future climate projections
title_fullStr Evaluation of uncertainty in a Maumee River Watershed Soil and Water Assessment Tool under current conditions and future climate projections
title_full_unstemmed Evaluation of uncertainty in a Maumee River Watershed Soil and Water Assessment Tool under current conditions and future climate projections
title_sort evaluation of uncertainty in a maumee river watershed soil and water assessment tool under current conditions and future climate projections
publisher The Ohio State University / OhioLINK
publishDate 2019
url http://rave.ohiolink.edu/etdc/view?acc_num=osu1555575109524802
work_keys_str_mv AT kujawahaleya evaluationofuncertaintyinamaumeeriverwatershedsoilandwaterassessmenttoolundercurrentconditionsandfutureclimateprojections
_version_ 1719455489997144064
spelling ndltd-OhioLink-oai-etd.ohiolink.edu-osu15555751095248022021-08-03T07:10:29Z Evaluation of uncertainty in a Maumee River Watershed Soil and Water Assessment Tool under current conditions and future climate projections Kujawa, Haley A. Environmental Science SWAT climate change uncertainty watershed modeling model validation Eutrophication threatens water quality and livelihoods globally. In Lake Erie, eutrophication and harmful algal bloom issues have persisted since the 1990s despite efforts to improve water quality. Phosphorus delivered to Lake Erie from the surrounding watersheds has been identified as the main limiting nutrient in harmful algal blooms, and therefore the focus of current water quality management is to reduce phosphorus inputs from Lake Erie’s watersheds. The current Great Lakes Water Quality Agreement calls for a 40% reduction in phosphorus input to Lake Erie compared to 2008 levels. The Maumee River Watershed covers much of the land draining to western Lake Erie and is therefore the largest contributor of phosphorus that drives harmful algal blooms. In the Maumee River Watershed, the majority of phosphorus comes from non-point source inputs in which a large portion is from agriculture. Water quality models can be used to run scenarios on how to effectively reduce phosphorus delivery to Lake Erie. To ensure accuracy of predictions and correct representation of the watershed, the watershed model is typically calibrated to an observed data site to ensure model outputs match reality. Outside of this site it is not known how well the model will perform, and data that could be used upstream of the calibration site is often discarded because it is often sparser and therefore deemed more uncertain. However, it could be argued this data, while more uncertain than calibration data, could be used to assess performance and inform modeling decisions. In the first chapter of this thesis, I compare a Maumee River Watershed Soil and Water Assessment Tool (SWAT) to multiple sites for measured discharge, total phosphorus, and dissolved reactive phosphorus during the calibration years (2005-2015) to assess prediction upstream of the calibration site. Discharge found performance improved with increasing watershed area and proximity to calibration gauge (NSE ranging from -1.38 to 0.78, median = 0.53). Total phosphorous (NSE ranging from -69 to 0.53, median = -0.7) and dissolved reactive phosphorus (NSE ranging from -0.38 to 0.39, median = 0.07) were more difficult to characterize because of the reduced frequency of measured data and the variability in results. Total phosphorous was overpredicted at more sites (70%) whereas dissolved reactive phosphorus was underpredicted at more sites (64%). This chapter finds while the model performs well at some locations upstream of the calibration gauge, this is not always the case and improvements could be made to the model and model inputs to improve accuracy.The Great Lakes Water Quality Agreement also calls for adaptive management to respond to a changing climate. Current water quality targets are based on stationarity in precipitation and temperature, but stationarity is not the case under climate change. Thus, predicted climate data from general circulation models (GCMs) can be used with water quality models to project anticipated changes in discharge and nutrient loadings. Since uncertainty from the climate data is large, a common method to characterize this uncertainty is to run the water quality model with an ensemble of GCMs. However, this method only captures the uncertainty in climate predictions. For the second chapter of this thesis, an ensemble of five watershed models of the Maumee River Watershed, configured in the Soil and Water Assessment Tool models were run with six GCMs. Predictions of future discharge and nutrient loadings by mid-century (2046-2065) were assessed and uncertainty in predictions driven by SWAT and GCMs was characterized. A signal to noise ratio was used to characterize ensemble agreement and analysis of variance to partition uncertainty between SWAT and GCM. Signal-to-noise results showed no clear agreement on the direction of change in future nutrient loadings or discharge. GCMs dictated the uncertainty in prediction of future discharge (96%) and total nitrogen (63%), while SWAT was more important in driving uncertainty in simulation of future phosphorus loadings (> 57%). These results demonstrate that use of one SWAT model may not be enough to characterize uncertainty in future nutrient loadings in this region, and improvements in both SWAT and GCMs are needed. 2019-08-27 English text The Ohio State University / OhioLINK http://rave.ohiolink.edu/etdc/view?acc_num=osu1555575109524802 http://rave.ohiolink.edu/etdc/view?acc_num=osu1555575109524802 unrestricted This thesis or dissertation is protected by copyright: all rights reserved. It may not be copied or redistributed beyond the terms of applicable copyright laws.

Similar Items