Effect of two flood routing methods and leaf area index on the calibration of a distributed hydrologic model

Water is a life source without any alternative for all living beings on the planet Earth. All humans, animals, plants and other life forms need it for their whole life time. Consequently, people devoleped hydrological projects for the optimal use of water due to a lot of purposes so as to make use of water. One of this useful project is finding discharge from basins. As a result of this, we want to know how to find best results and how to calibrate models. There are a lot of benefits of finding discharge on any water sources such as we can prevent floods, build water structures according to this data. Also we can use as an irrigation area. Therefore, we should always focus on finding new methods and new calibration techniques in order to find optimum discharge and other values of water sources. Otherwise, we can not benefit from them efficiently and animals,plants and other life forms may suffer. In developed countries, they study on this everyday and build water structures based on new values. In the world, there are plenty of algorithms, methods, objective functions and other parameters that adjust flow on a basin. This study inform us that methods, objective functions and leaf area index. We can use these methods, objective functions other than hydrological area. Because of this, these are all relate to mathematical equations. Thus, we may see these methods any other area's research. As a result of this, we should check every method, objective function and use each of it in many variations in order to get the best result from calibrating the model. Also we save time and money when we use correct method, objective function and leaf area index. When we use correct method, objective function and leaf area index we can find optimum result in a short time. In this study, we aim on effects of methods, objective functions and leaf area index on the calibration od a distributed hydrologic model. In the Chapter 1, purpose of thesis is expLAIned. Furthermore, detailed informations were given about research. It is expLAIned that we aim for the best optimization for hydrological models. We use mHM program so as to find good results. Furthermore, Notepad++ and Excel programs are used to get the values of daily discharges and dds results. In the Chapter 2, our research project based on Skjern basin in Denmark. So the basin area explained and gives us detailed informations. All calculations stand on Skjern basin area. In the Chapter 3, methods, oblective functions and leaf area index are introduced. Dynamically Dimensioned Search (DDS) algorithm used on this research due to its efficiency. Other algorithms such as MCMC, Simulated Annealing and SCE are not used. There are two routing methods which are Muskingum routing method and Adaptive time step routing method. NSE, Ln NSE, KGE and SSE are used as objective functions. Also we give leaf area index (LAI) value whether 0 or 1 in order to turn it off and on. When we give LAI input 1 it means it is on. When we give LAI input 0 than it is gonna be off. We have 2 methods, 4 objective functions and 2 LAI input values. As a result of this, 16 different data are collected. In the Chapter 4 shows us results of this calibration of the model. Dynamically Dimensioned Search is always used. However, as we mention before there are Muskingum and Adaptive routing methods, NSE, Ln NSE, KGE and SSE objective functions and lastly 2 LAI input values which are 0 or 1. We obtained data for Muskingum method, NSE objective function and LAI input 1, Muskingum method, NSE objective function and LAI input 0, Adaptive method, NSE objective function and LAI input 1, Adaptive method, NSE objective function and LAI input 0, Muskingum method, Ln NSE objective function and LAI input 1, Muskingum method, Ln NSE objective function and LAI input 0, Adaptive method, Ln NSE objective function and LAI input 1, Adaptive method, Ln NSE objective function and LAI input 0, Muskingum method, KGE objective function and LAI input 1, Muskingum method, KGE objective function and LAI input 0, Adaptive method, KGE objective function and LAI input 1, Adaptive method, KGE objective function and LAI input 0, Muskingum method, SSE objective function and LAI input 1, Muskingum method, SSE objective function and LAI input 0, Adaptive method, SSE objective function and LAI input 1, Adaptive method, SSE objective function and LAI input 0. We collected data each of this then outputs opened on Notepad++ then transferred into Excel in order to find total error and optimum iteration number. In the Chapter 5, it is emphasized that we should use mHM to calibrate hydrological model. In the lights of this research when we use Dynamically Dimensioned Search as an algorithm for mHM program we can take iteration number as 20. We can say that this number is the optimum number due to results of the all methods, objective functions and leaf area index. When we drop iteration number from 100 to 20 we save up to 5 times in time and save money. Because when we use 100 iteration number computer process in approximately 3 hours. However, when we take iteration number as 20 it last just 35 minutes. Therefore, we should check basin area for this and run calibration for model in order to find optimum result. It is recommended that if a model is calibrated by using mHM programme iteration number should not take any higher values than 20 or some case 30.