Kyle Hafliger

Kyle Hafliger is a Master of Engineering student in the Civil Engineering program at the University of North Dakota, Grand Forks. He holds a Bachelor of Science degree in Civil Engineering from the North Dakota State University. His present research is focused on different techniques for accessing changes in flooding trends in the Upper Midwest and causes of flooding in the selected rivers with changes in trends.

E-mail:
kyle.hafliger@und.edu
Phone: 701-371-8609

Techniques of Assessing Changes in River Flooding Patterns in the Upper Midwest

Fellow: Kyle T. Hafliger

Advisor: Yeo Howe Lim, P.E., Associate Professor, Department of Civil Engineering, University of North Dakota.

Degree Progress: M.E. in Civil Engineering expected graduation in December 2011

Research:

Many rivers in the upper Midwest have generally had an increasing trend in flooding over the past forty years. It is casually observed that since 1993 the Upper Midwest has been in a wet cycle with above average precipitation which usually leads to more flooding. The purpose of this research is to study changes in the flooding patterns of rivers or streams in the Upper Midwest. Annual peak flow series of various rivers are to be scrutinized for trends using different statistical methods. These methods include: the two tailed t-test, the moving average, the Mann-Whitney test, and the Mann-Kendall test. Based on previous studies done on trend testing and the results that have been found in this study, the Mann-Kendall test is found to be a good fit. The two time periods chosen for comparison in flooding trends was from the period prior to 1970 and the period from 1971 to present. Next in the research, the daily flow data was analyzed for each of the rivers that had a significant change in the annual peak flows to look at the flood changes in more detail. Wavelet plots are used in this study to investigate the changes of daily flow patterns. A wavelet plot is a plot of the power frequency of the variable measured versus time. The wavelet plot shows how often patterns can occur with different time intervals. When the wavelet plot was analyzed, it showed that the most frequent time pattern for flooding was approximately one year. This makes sense, because that is the recurrent time period for the annual spring runoff, which usually is when the peak flows happen. Another noted time period that had a higher wavelet powers was about three years. About once every three years, an El Nino or a La Nina weather pattern usually affects the United States, which can have an effect on the climate. Also in this study, other reasons that caused the increase in flooding are investigated. It could be from changes in amount of precipitation, temperature changes, or changes in evaporation. Some other anthropogenic factors things that can affect the flooding could be different land uses in the basin, which affects the runoff into the river. In this study each selected river that had a flooding change will have a reason listed for what caused the flooding.

The present research is in continuation of the work proposed for the 2011 NDWRRI Fellowship.

Project Objectives:

Progress:

Initially, a couple rivers in the Upper Midwest were selected as of right now. These rivers were tested for trend analysis for the peak annual flows. Several different statistical methods were investigated. Initially, the two tailed t-test was done, but it doesn't show the trend for the data as a whole, just one set of data compared to another. The Mann-Kendall test was chosen since it shows the trend for the data as a whole. The streams are assumed to have an increasing trend for p values less then 0.05. As soon as the wavelet analysis method for the daily data is analyzed, several other streams can be analyzed for the annual peak flow and daily data. In the future, the flooding causes will be investigated.

Significance:

The Upper Midwest has generally been in a wet cycle since 1993. This has caused increased flooding rivers and streams in many areas. This study will select certain rivers that have undergone a change in flooding trends and analyze them for why the flooding trend might have increased. Knowing which rivers have an increase in flooding trends and which ones do not will be beneficial in the long run. In the rivers that are found to have an increase in flooding, actions then can be made to lessen the flooding such as building dams or diversions.

Dr Yeo Howe Lim

Advisor: Yeo Howe Lim, P.E.

Associate Professor, Department of Civil
Engineering, University of North Dakota

Phone: (701) 777-3998

Fax: (701) 777-3782