Hasin Munna

Hasin Shahad Munna is a graduate student completing a MS Degree in Civil Engineering at the University of North Dakota. He graduated with a Bachelor of Science Degree in Civil Engineering from Bangladesh University of Engineering and Technology (BUET), Bangladesh in 2009. His current research is focused on flood risk assessments of various scenarios for Devils Lake under GCM downscaling simulations using a coupled hydro-climatic model incorporating recent advances in lake evaporation estimations.

Email: hasin.munna@und.edu
Phone: 701-610-1645


Flood Risk Assessments of Various Scenarios for Devils Lake under GCM Downscaling Simulations Using a Coupled Hydro-Climatic Model Incorporating Recent Advances in Lake Evaporation Estimations

Fellow: Hasin Shahad Munna

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

Degree Progress: M.S. in Civil Engineering. (Expected Graduation in Fall 2011)


Located in the Ramsey and Benson County, Devils Lake is a closed basin that receives surface runoff from 3,810 square miles of drainage basin (USGS). The uncontrolled growth of the lake has been an alarming issue for North Dakota for the last few years. Over the past 15 years, the North Dakota State Water Commission (NDSWC), in collaboration with the Devils Lake Basin Joint Water Resource Board, has implemented numerous flood mitigation projects for Devils Lake, costing more than $500 million, to raise levees, improve roads, and build an outlet. The rising water has destroyed hundreds of homes and businesses, inundated thousands of acres of productive farmland, and displaced nearly 1,000 people, about 15% of the population. The on-going lake flooding is both unique and complex which warrants much greater research efforts especially in the water resources engineering and management perspectives.


The proposed research consists of 3 basic parts.

  1. Modification and recalibration of different models to estimate the evaporation time series:
    A total of eight radiation based evaporation models have been analyzed to verify the best model in Devils Lake region. Among these the Hargreaves method, Doorenbos and Pruitt method and Makkink method showed relatively close matches with the observed values from nearby gauging stations. A simplified Penman equation was also examined to check its efficiency in estimating the evaporation. The research will focus on recalibrating the proposed methods (Makkink, Doorenbos and Pruitt, Hargreaves and simplified version of Penman equation) to obtain better estimations.
  2. Applying the simulated GCM based temperature and precipitation time series to the calibrated and coupled hydro-climatic model to generate future lake levels:
    In this part the generated synthetic daily temperature and precipitation time series will be applied to the coupled model. Part of the task is to identify a common compatible format to import the synthetic datasets of numerous different scenarios into the coupled model. The HEC model uses a powerful database tool called HEC-DSSVue to import time series parameters. DSSVue can import data in ASCII format which can be converted into .dss format by using utility plug-ins. A properly oriented ASCII format will be designated for the interchange of data. After identifying the proper way for exchanging data, simulations will be run which will give us future simulated water levels of Devils Lake at different selected time periods (2020, 2050) under varying climatic conditions.
  3. Frequency analysis of the synthetic peak stages of Devils Lake:
    There are different methods for fitting the probability distribution of peak values some of which are 2-parameter lognormal (LN2), 3-parameter lognormal (LN3), Gumbel distribution, general extreme value (GEV), Log-Pearson 3 (LP3) and Pareto distributions. For estimating PD for peak flood frequency analysis, the Log-Pearson Type III (LP3) distribution is recommended (Bulletin 17B) according to the U.S. Water Advisory Committee on Water Data (1982). Peak stage analysis will be done using the recommended method and also by other popular methods which will give a comparative measure of the goodness of fit. The distribution method with the best fit and also the recommended LP3 method will be applied in the frequency analysis of the simulated data to estimate peak stages for different return periods.

The research focuses in calibrating a coupled rainfall-runoff model and a reservoir model for the Devils Lake using data from both ground gage stations and NASA satellite observations. The purpose is to determine the feasibility of using spatially distributed GCM data with a well calibrated hydro-climatic model to predict the probable flood severities. A temperature-based evaporation prediction model is also developed to simulate the outflow from the terminal lake. Future hydrology of the basin and lake levels of Devils Lake are simulated using the weather samples obtained from several downscaled GCM runs under varying scenarios due to anthropogenic modifications and the resulting composition of the atmosphere. 100 traces of future water levels of the Devils Lake have been generated using the predicted temperature and precipitation by the GCMs. The synthetic traces show a downward trend in water levels for a 30 year simulation period. The annual peak series of the synthetic traces (both stage and volume) are sorted and analyzed to obtain the probabilities and return periods of extreme flood events. The Bulletin 17B recommended LP3 method along with Gaussian/Normal, Lognormal and Gamma/Pearson type 3 distributions are applied for comparison purposes. The Log-Normal probability distribution, in both cases (stage and volume) provided with better fits. Water levels of closed-basin lakes are usually characterized by high serial dependence. The lake is already in its highest peak of the recorded history (1454 ft, spring, 2011), which is set as the base condition for the simulation of future water levels. These circumstances require the probability to be calculated both in a conditional and unconditional basis. Moreover, the probability is calculated both in terms of stage and volume, and probabilities of water levels being 1456, 1458 and 1460 ft have been reported. Considering the stage of the lake, the conditional probabilities are 0.008, 0.003 and .001 percent and the unconditional probabilities are 1.12, 0.44 and 0.189 percent respectively. By converting the stages into volumes, the conditional probabilities are 0.004, 0.002 and 0.0009 percent and the unconditional probabilities are 1.5, 0.68 and 0.31 percent respectively. These lower probability values indicate a lower chance of spilling into the Sheyenne River in the near future based on GCM predictions.


Through analysis and recalibration, the best model for estimating ET will be selected. It is expected that the use of this model will allow better simulation of the water levels of the lake. Future ETs for simulation purposes will be estimated and used in the coupled hydrologic model.

The simulations using downscaled GCM based precipitation and temperature will generate the future probable water levels of the Devils Lake for varying climatic conditions.
The frequency distribution of the peak stages using different methods will suggest the best method to be used for this particular study area. Bulletin 17B recommended method will also be analyzed. The comparison will give us an idea about the goodness of the fit.

The frequency analysis of the synthetic peak stages can be a valuable decision making tool in estimating peak stages for different return periods. The analysis can be very helpful in taking critical decisions and designing flood mitigation structures. The analysis will also determine the risk associated to flooding in near future and will work as a guide for the city planners


Lim, Y.H., Munna, H.S., Zhang, X., Kirilenko, A. and Teng, W. (In preparation). “Flood Level Exceedance Probabilities of a Terminal Lake under Down-scaled GCM Predictions Using a Coupled Hydro-climatic Model”.

Conference Presentations and Proceedings:
Munna, H.S. (2012, May). “Development of a Temperature-based Model to Simulate Evaporative Losses over Water Bodies in Cold Regions”, World Environmental and Water Resources Congress, ASCE-EWRI, Albuquerque, New Mexico.

Lim, Y. H., Zhang, X., Kirilenko, A., Teng, W. and Munna, H. S. (2012, May).  “Flood Frequency Analysis of Devils Lake under Current and Projected Future Climates Utilizing HEC Hydrologic Models, NASA Satellite Observations and Downscaled GCM Simulations”, World Env. and Water Res. Congress, ASCE-EWRI, Albuquerque, New Mexico.

Munna, H.S. and Lim, Y. H. (2012, May). “An Efficient Method for Predicting Evaporative Losses over Water Bodies in Cold Regions using a Temperature-Based Model”, Poster Paper, World Env. and Water Res. Congress, ASCE-EWRI, Albuquerque, New Mexico.

Munna, H.S. and Lim, Y. H. (2010, Dec). “Coupling of HEC-HMS and HEC-ResSim in Modeling the Fluctuation of Water Level in Devils Lake Using Heterogeneous Data”, Poster Paper, Abstract H43F-1324 presented at 2010 Fall Meeting, American Geophysical Union, San Francisco, California. 

Seminar Presentation:
Seminar presentation on “Water Level Quantiles of a Terminal Lake under Future Climate Change Conditions” (2012, April) at the “Terrestrial Water Cycle Seminar”, NASA’s Goddard Space Flight Center, Greenbelt, Maryland.


Dr. Howe Lim

Advisor: Howe Lim
Assistant Professor of Civil Engineering
University North Dakota