Dissertation
An investigation of Iowa’s riverine phosphorus loads and statewide phosphorus budget
University of Iowa
Doctor of Philosophy (PhD), University of Iowa
Spring 2022
DOI: 10.25820/etd.006583
Abstract
Phosphorus is a chemical of foremost concern within the state of Iowa. While it has an essential role in biological functions, it also acts as a pollutant after entering surface water. Phosphorus plays a significant part in Iowa’s economy through its role in many agricultural, municipal, and industrial activities. It is also readily present in the state’s soils. Consequently, Iowa struggles with its presence in the state’s water. Elevated levels of phosphorus are commonplace, and eutrophication is a considerable concern in the state’s waters and waterbodies located downstream. In response, numerous stakeholders have invested significant time and effort into reducing its aquatic presence—primarily by implementing practices that try to prevent P from reaching Iowa’s waters. The Iowa Nutrient Reduction Strategy has set a target of reducing P loads exported through the state’s rivers by 45%. Several challenges exist in estimating these loads and, consequently, determining if progress is being made towards this reduction goal. Waterborne phosphorus must be measured in a laboratory setting; it cannot be quantified on-site. Riverine phosphorus is dynamic and often prone to high concentrations that contribute to substantial positive skew when flux is estimated. Quantifying statewide loads by simply measuring phosphorus in-situ is infeasible. This is due to both limitations with field-based phosphorus measurements and the uncertainties that arise between discrete phosphorus samples.
To overcome these challenges, I developed several statistical models that predicted riverine phosphorus concentrations. This process required assembling and evaluating surface water data collected at 16 sites, each located along a major river near Iowa’s border. Data were acquired from the Iowa Department of Natural Resource’s AQuIA database and the United States Geological Survey’s Water Quality Portal. The resulting dataset was a robust collection of phosphorus samples (specifically total phosphorus, orthophosphate, and particulate phosphorus) alongside coincident streamflow. This dataset primarily consisted of monthly samples collected from 1998 to the present. Most phosphorus concentrations followed lognormal distributions, with most values ranging from <0.01 to 1 mg/L. Larger concentrations did sometimes occur, usually in the summer for particulate phosphorus and winter or early spring for orthophosphate. Likewise, concentrations varied across sites, with statistically significant differences present for both the dissolved and particulate forms. The dissolved and particulate forms were not correlated, so I modeled these as separate dependent variables.
For particulate phosphorus, turbidity-based surrogacy models were the most appropriate. These models used high-resolution in-situ turbidity measurements to predict particulate phosphorus concentrations. The Weighted Regressions on Time, Discharge, and Season statistical package was most suitable for modeling orthophosphate as well as particulate phosphorus when turbidity data were unavailable. These types of models incorporate daily streamflow values and seasonal metrics to predict phosphorus concentrations. The results were further improved by deploying a Kalman filter which adjusted the models’ estimates to better match observed phosphorus data.
These models estimated daily phosphorus concentrations from 1998 to 2020, yielding the most detailed phosphorus loads to date. An extensive turbidimeter network, the data of which are made publicly available through the Iowa Water Quality Information System, provided much of the data needed for model implementation. Daily mean streamflow values were used to estimate daily phosphorus loads at each of 16 sites. Annual statewide loads ranged from 5 to 40 million kg over the simulated period, with a yearly average near 20 million kg. Large year-to-year variations in loads were observed, and changes were consistent with precipitation. Iowa was found to have an overall phosphorus yield of 2.0 kg/ha/yr. for the period of record.
In 2021, I conducted a fieldwork campaign that collected new phosphorus data in Eastern Iowa. The existing samples were found to lack measurements taken during periods of high flow, so this campaign focused on data acquisition during periods of elevated streamflow. These newly collected data validated the use of the previously described models under the full range of hydrologic conditions.
I also quantified phosphorus inputs and outputs to construct a statewide phosphorus budget. The budget included phosphorus sources from fertilizer and manure application, as well as human and industrial sources. It also included harvested phosphorus and phosphorus consumed via livestock grazing, in addition to the phosphorus lost through Iowa’s rivers. Each of these components was quantified from 1998 to 2019 to produce annual P estimates, tracking the yearly net flux of P in Iowa. The fertilizer and manure components dominated P inputs, while harvesting dominated the outputs. In most years, phosphorus outputs were greater than inputs, leading to a net negative budget. Additionally, the riverine phosphorus component did not correlate strongly with any of the budget’s other parts.
Soil sampling measurements were also used to quantify the phosphorus present within Iowa surficial soil. The Iowa State-Wide Trace Element Soil Sampling Project measured phosphorus in over 530 locations within the first 2 ft of Iowa’s surficial soils. This calculation yielded a value of 81.5 billion kg—far greater than any of the individual components of the phosphorus budget.
The findings of this study produced novel P estimates for the various watersheds of Iowa. At the same time, its methods can act as a template for other locales seeking to evaluate phosphorus reduction progress. These results supplement the work conducted by Iowa’s stakeholders to protect the state’s surface water resources against eutrophication.
Details
- Title: Subtitle
- An investigation of Iowa’s riverine phosphorus loads and statewide phosphorus budget
- Creators
- Elliot S Anderson
- Contributors
- Larry Weber (Advisor)Keith Schilling (Advisor)Christopher Jones (Committee Member)Gabriele Villarini (Committee Member)Corey Markfort (Committee Member)Jessica Meyer (Committee Member)
- Resource Type
- Dissertation
- Degree Awarded
- Doctor of Philosophy (PhD), University of Iowa
- Degree in
- Civil and Environmental Engineering
- Date degree season
- Spring 2022
- Publisher
- University of Iowa
- DOI
- 10.25820/etd.006583
- Number of pages
- xxi, 252 pages
- Copyright
- Copyright 2022 Elliot S Anderson
- Language
- English
- Description illustrations
- color illustrations, color maps
- Description bibliographic
- Includes bibliographical references (pages 196-211).
- Public Abstract (ETD)
One of Iowa’s most needed chemicals is also one of its most harmful. Large amounts of phosphorus are present in Iowa’s soil, and it is used extensively in fertilizer needed to help the state’s crops grow. However, when phosphorus ends up in Iowa’s water, it causes algae to bloom. These blooms make waterbodies inhospitable for fish and other aquatic life, and many often die as a result. Certain types of algae can also poison any humans or animals that encounter them. The worst-case scenario occurs when communities rely on a water body as a drinking source or site for recreation, and due to algae, it is no longer available.
Over the past several decades, Iowa has formulated a plan whereby its stakeholders work together to reduce aquatic phosphorus. This plan implements several practices that better treat rainwater runoff and wastewater before reaching Iowa’s rivers. A problem arises when people have tried to assess the progress of this plan: phosphorus is difficult and expensive to measure. Water sampling is only part of the effort needed to estimate statewide phosphorus totals, and determining trends can be challenging.
This study resolves this problem by monitoring turbidity (how murky the water is). Turbidity is much easier to measure than phosphorus, and I developed statistical models that use it to predict phosphorus across Iowa. I also built additional models that use river flow and seasonal information to predict phosphorus when turbidity is unavailable. With these models, I estimated the amount of phosphorus being exported by Iowa’s rivers—providing insight into the progress of the state’s phosphorus reduction efforts. These findings will shed light on which phosphorus-reducing practices work best for specific regions of Iowa. The state will benefit when it knows precisely how it can best keep phosphorus, and therefore algae, at bay.
- Academic Unit
- Civil and Environmental Engineering
- Record Identifier
- 9984271356002771
Metrics
7 File views/ downloads
65 Record Views