UNDERSTANDING THE RELATIONSHIP BETWEEN CLIMATE AND CONCENTRATION DISCHARGE: AGRICULTURAL NITRATE IN THE CLEAR CREEK WATERSHED

Abstract

 

The use of agricultural nitrate sources is a growing concern in Midwestern United States. Since the late 1980s, studies have shown that seasonal hypoxic zone formations at the Gulf of Mexico is mostly the result of high levels of nutrients being exported from Midwestern streams. The main source of nutrients in Midwestern streams is due to increases use in agricultural fertilizers.   This study examines the concentration-discharge (C-Q) signature in streams draining agricultural lands using Time series analysis, frequency analysis and model simulation for nitrate-Nitrogen. The study uses the Soil and Water Assessment tool (SWAT), a watershed scale model to predict the impact of land practices on an agricultural watershed using historical observations and future projection data. These include five years observational field data and the InterGovernmental Panel for Climate Change (IPCC) model 3 analysis data for a fifty-year climate projection. The overarching goal of the study is to understand observed long term increases in historical nitrate stream concentration, evaluate management practices associated with an agriculture watershed in the midwest and to evaluate the potential for climate change to alter the future nitrate and nitrate loads in the watershed. The study area is the Clear Creek watershed (CCW) located in east-central Iowa. The CCW joins the Iowa River in Coralville, Iowa. The watershed is representative of many Midwestern watersheds with humid-continental climate with predominantly agricultural land use. The SWAT model is the basic catchment model used in this research. Calibration with SWAT shows reasonable result (NSE=³0.5) based on the            Nash-Sutcliff criterion (adapted from Santhi et al., 2001).

iii

 

 

In this study, modeling, time series and data analysis tools were used to evaluate historical nitrate conditions in an agricultural watershed subject to high nutrient loading, and how climate projection data may be applied to evaluate best management practices in the future. Time series (Spectral and Frequency) analysis and Multivariate (Principal Component Analysis) analysis suggests seasonal stream discharge is a dominant driver of nitrate concentration at the watershed outlet. Analysis of the trends in the concentration-discharge phase plain show a seasonal increase in nitrate during the wet season and nitrate utilization during the growing or dry season (late May through August). The major sources/sinks of nitrate include plant nutrient uptake and seasonal fertilizer application

Table of Contents

LIST OF FIGURES ………………………………………………………………………………………………………………… 7

LIST OF TABLES ………………………………………………………………………………………………………………….. 9

ACKNOWLEDGMENTS …………………………………………………………………………………………………….. 10

CHAPTER 1  |………………………………………………………………………………………………………………… 1

INTRODUCTION AND BACKGROUND………………………………………………………………………… 1

1.1 PROBLEM STATEMENT…………………………………………………………………………………………………… 1

1.2  RESEARCH OBJECTIVE/HYPOTHESIS………………………………………………………………………………… 3

CHAPTER   2  |……………………………………………………………………………………………………………….. 4

STUDY      AREA  AND  MODEL  DESCRIPTION……………………………………………………………. 4

2.1 STUDY AREA……………………………………………………………………………………………………………….. 4

2.1.1 Past Human Impacts……………………………………………………………………………………………. 6

2.1.2 Climate……………………………………………………………………………………………………………… 7

2.1.3  Topography and Geology………………………………………………………………………………………. 8

2.2  MODEL…………………………………………………………………………………………………………………….. 11

2.2.1  SWAT Description and Model Theory…………………………………………………………………… 11

2.3  LITERATURE REVIEW………………………………………………………………………………………………….. 12

2.3.1  Using SWAT for Nitrogen loses and C/Q dynamics………………………………………………….. 14

2.3.2  Long-term Climate Models and studies………………………………………………………………….. 15

CHAPTER   3  |……………………………………………………………………………………………………………… 17

INPUT DATA AND METHODS……………………………………………………………………………………. 17

3.1 INPUT DATA………………………………………………………………………………………………………………. 17

3.1.1  SWAT Inputs…………………………………………………………………………………………………… 17

3.1.2  NLDAS-2, USGS and SWAT data Sources…………………………………………………………….. 19

3.2  METHODS USED IN SWAT AND IPCC DATA INPUTS……………………………………………………….. 19

3.2.1 Water Balance…………………………………………………………………………………………………… 19

3.2.2 Time Series Analysis…………………………………………………………………………………………… 20

3.2.3 Bias Correction and Downscaling………………………………………………………………………….. 20

3.2.4  Quantile Mapping…………………………………………………………………………………………….. 22

CHAPTER 4  |……………………………………………………………………………………………………………….. 25

MODEL IMPLEMENTATION AND CALIBRATION……………………………………………………… 25

4.1  MODEL IMPLEMENTATION………………………………………………………………………………………….. 25

4.2  MODEL CALIBRATION/VALIDATION……………………………………………………………………………… 26

4.2.1  Default Scenario……………………………………………………………………………………………….. 26

4.2.2 Model Calibration/Validation……………………………………………………………………………….. 27

CHAPTER 5 | ……………………………………………………….^v……………………………………………………… 32

RESULT AND DISCUSSIONS………………………………………………………………………………………. 32

5.1 INTRODUCTION………………………………………………………………………………………………………….. 32

5.2  TIME SERIES DATA AND SIMULATION 2011-2016…………………………………………………………….. 33

5.2.1  Time series Analysis………………………………………………………………………………………….. 35

5.2.2  C/Q relationship……………………………………………………………………………………………….. 45

5.2.3  Loading…………………………………………………………………………………………………………… 49

5.3  CLIMATE CHANGE PROJECTION……………………………………………………………………………………. 53

5.3.1  Concentration Discharge Relationship (2018 through 2051)……………………………………….. 54

CHAPTER 6  | CONCLUSION……………………………………………………………………………………….. 56

BIBLIOGRAPHY…………………………………………………………………………………………………………… 58

Chapter 1                              |

Introduction and Background

 

1.1 Problem Statement

Runoff from agricultural land is a major concern for stream and groundwater quality in the mid-western agricultural belt. In particular, high levels of pollutants such as nitrate, organic carbons, phosphorous etc. are found in most mid-western streams. Given the intensity of agriculture and the large fraction of land in agriculture in the upper mid-west, high concentrations of contaminants found in these streams does affect downstream water quality in tributaries and main stream of the Mississippi (see Figure 1). Studies have shown that the development of hypoxic conditions at the Gulf of Mexico is largely contributed to the nutrients exports from mid-western streams (e.g. Rabalais et al., 1996; Goolsby et al., 1999). About 25% of the nitrate deliver by the Mississippi river to the Gulf of Mexico comes from the State of Iowa, despite Iowa only occupies less than 5% of the Mississippi drainage basin (Schilling and Libra, 2000). Nitrate in particular is of major interest since much of the nitrogen in the stream is transformed into nitrate (the main contaminant

 

 

 

 

 

Figure 1. 1: Vulnerability of Midwestern United State water quality and the amount of nitrogen in put alone with the CCW shown with the black circle (Modified from: Link )

 

in surface and ground water) over the mid-west. The goal of this work is to understand the main drivers of nitrate contamination in the past and the present for the future using available observations, modeling and data analysis tools. This study uses a typical mid-western watershed which was instrumented by the IIHR-Hydroscience & Engineering at the University of Iowa, the Iowa

Water Quality Information System (IWQIS) and the Intensively Managed Landscape Critical Zone Observatory (IML-CZO). The CCW is a small watershed which sits in a high nitrogen input and aquifer vulnerability.

 

 

 

 

 

1.2 Research Objective/Hypothesis

The main objective of this research is to understand the major cause(s) of the continuous increase in nitrate concentration using 5 years measured nitrate concentration. Studies have shown that high accumulation of nitrate concentration at the Gulf of Mexico is mostly attributed to nitrate exported from Midwestern streams (e.g. Schilling Zhang, 2004; Royeretal, 2006) like the CCW. A five-year analysis (2012 through 2016) used the SWAT model to simulate and calibrate flow and nitrate transport in an agricultural setting. Future projections were made using IPCC climate forcing to estimate   potential future nitrate trends for study site, typical of mid-western agriculture. For both future projections and past observations, we analyze the effect of seasonal, annual and plant uptake on the transport of nitrate at the CCW. In addition, we study trends in nitrate concentration, the main drivers of nitrate export and the effect of climate change at the watershed scale. However, one limitation for the future projections is that we assume that the current management practices persist (or experiences only small changes) unto the future. To achieve the overarching objective of this work, we tested several concerns including the main driver of nitrate concentration (Can we capture the main driver using concentration discharge analysis), the accuracy of the SWAT model compared to the observed data as well as statistical analysis of both the model and the data and lastly the extent to which we can represent future processes based on projected data. Comparison of observed data with the model were found to have acceptable errors (not more than five percent) for the means, variance, standard deviation and skewness if all processes of the CCW are to be represented. Furthermore, the high level of agricultural fertilizer uses at the CCW coinciding with the seasonal flow patterns, the main driver of nitrate concentration was found to be a direct function of both parameter (flow discharge and management practices). Lastly, we propose that downscaling and bias correction of the IPCC projected climate data can improve the analysis of future nitrate trends.

UNDERSTANDING THE RELATIONSHIP BETWEEN CLIMATE AND CONCENTRATION DISCHARGE: AGRICULTURAL NITRATE IN THE CLEAR CREEK WATERSHED

Sharing is caring!