Application of HEC-HMS modelling on River Storån,Model evaluation and analysis of the processes by using soil moisture accounting loss method

Abstract

River Storån, a major tributary of Lake Bolmen which is a source of fresh water to Skåne county of Sweden is found to alter the color of water in the lake due to the presence of natural organic matter. Hence it is of utmost importance to study the river to understand the dynamics of generation and flow of water from the river to the lake. This is done by application of a hydrological modelling system called HEC-HMS (version 4.3) developed by Hydrologic Engineering Centre, US Army corps. River Storån is modelled using soil moisture accounting loss method, a method chosen for simulations run for a long period of time. During the application, data collected are assembled and input into the model components of HEC-HMS. Later the model is calibrated from 2005 to 2010 followed by optimization and validation from 2011 to August 2019.
The results obtained showed that the model underestimated the observed flow over the years and the reasons for underestimation was attributed to three factors, namely the climatic conditions, geological condition and topological condition. The temperature being an important climatic condition influenced processes like snowmelt and evapotranspiration, due to its inaccuracy it failed to recognize the type of precipitation specially during winter and spring. This led to the model generating low flow whenever the temperature went below 0. Further the generalization with respect to soil characteristics and depth led to quick response of the model to the precipitation which resulted in the simulated flow being more dependent on baseflow than on surface flow. Lastly, the exclusion of wetlands and small water bodies across the basin disregarded its impact on the process of evapotranspiration over the years.
Through the model evaluation it was found that the RSR, NSE and Pbias value of both calibration and validation was found be same i.e., 0.9, 0.2 and -1% respectively. The PEPF was found to be 18% and 16% for calibration and validation and PEV was found to be 1% and -3% respectively. Although the model evaluation does not meet the expected value it is found to be satisfactory as the model successfully generates total volume, peak discharge and date of peak discharge very close to the expected value. This was followed by a discussion on the limitations of the model and recommendations to overcome it.
The hydrological model HEC-HMS application to River Storån did not give the expected result but was found to work well for the areas which is prone to snowfall and heavy precipitation all through the year. The model is user friendly and is found compatible with Arc GIS which makes the parameterization easy. Further research and study are highly recommended to upgrade the model for better efficiency of the model.

Popular Abstract

Lake Bolmen is the main source of drinking water to the central part of Scania, Sweden. Due to increase in the natural organic matter the water in the lake is turning brown. This is attributed to the inflow from its main tributary River Storån. The current project focuses on River Storån. Here a hydrological model called HEC-HMS developed by US Army Corps is applied on the river to simulate a flow from 2005 to August 2019. Further the model is evaluated wherein the processes that could possibly have an impact on the results are analysed. Lastly a discussion is done on the limitations of the model and recommendations are suggested to overcome it.
The hydrological cycle (water cycle) is a very interesting and complex process of nature. It involves several processes and parameters which is difficult to mimic by any hydrological model. Hence in HEC-HMS a loss method called Soil Moisture Accounting (SMA) method is chosen which was found to represent the actual water cycle very closely. This is a method preferred for long term simulation as it works in conjunction with other losses above soil surface.
The model is calibrated from the year 2005 to 2010 and validated from 2011 to August 2019. The resulting simulated flow is evaluated against the observed flow value obtained from SMHI which showed an underestimation of observed flow during calibration and a slight over estimation during validation. The model tends to underestimate the flow specially during winter and spring seasons in contrast to reality. Further the model was found to have a quick response to the precipitation, i.e., the model generated flow whenever there was any precipitation with very little time gap. This was attributed to three main conditions which is said to have major influence on the stream flow namely, climatic condition, geological condition and topographical condition.
Temperature is the most important climatic factor which influences processes like snowmelt and evapotranspiration. Due to the inaccurate temperature data the model failed to recognise the type of precipitation and as a result whenever the temperature went below 0°C the model generated bare minimum flow. Similarly, the evapotranspiration value used didn’t allow the possibility of variation of any sort over the years and lead to the underestimation of loss due to evaporation. Further the assumptions made regarding geological aspect like the depth and composition of soil lead to the quick response of the model to the precipitation falling resulting in the dependency of simulated flow more on baseflow than on overland flow. During modelling several wetlands, small water bodies were not taken into consideration and it is believed to have affected the result as well. The model evaluation done using indices like NSE, RSR, PEV, PEPF although showed somewhat satisfactory result, it did successfully recognise the total volume and date and volume of peak discharge very close to the observed value. The limitations and recommendations about the usage of the model were mainly regarding the simplification of large basin into workable subbasins, reduction of generalisation with respect to parameters, collection of accurate data etc. The model however was found to be user friendly which was capable of imitating the actual water cycle to a great extent even for an area like the present which has high rainfall and snowfall during most part of the year. This model did successfully shed a light on the processes involved in the River Storån generating an inflow to Lake Bolmen. Further research and study are recommended with regards to the model HEC-HMS to improve its efficiency and applicability.

• author: Makki Vijayaprakash, Pooja
• supervisor: Magnus Persson; Clemens Klante
• organization: Division of Water Resources Engineering
• course: VVRM01 20191
• year: 2021
• type: H2 - Master's Degree (Two Years)
• subject: Technology and Engineering
• keywords: River Storån, HEC-HMS, SMA, streamflow, baseflow, evapotranspiration
• publication/series: TVVR20/5021
• report number: 20/5021
• ISSN: 1101-9824
• language: English
• additional info: Examiner: Magnus Larson
• id: 9035455
• date added to LUP: 2021-01-12 09:38:57
• date last changed: 2021-01-12 09:38:57