Effect of dynamic pet scaling with LAI and aspect on the spatial performance of a distributed hydrologic model

Abstract

There are many valuable materials in world but non of them is as vital as water. Humankind has wanted to control water throughout history. First calendars were invented to predict dry and wet seasons. Floods of rivers were calculated, and irrigation systems designed. Astronomy was used to monitor the movements of Sun, Moon, and Earth to predict tides. Therefore, desire to control water is important reason to lead people to build megastructures. Today, huge dams were built for irrigation, flood control, water supply and energy production for cities. With the advancement of technology, opportinutities and development in observation system are increased. For example, Global Navigation Satellite Sysytem (GNSS), photogrammetry, Light Detection and Ranging (LIDAR) and Radio Detecting and Ranging (RADAR), Geographic Information System (GIS), and remote sensing. Additionally, surveying, and geological techniques are developed. Therefore, to improve model many different parameters are taken consideration. For example, temperature, humidity, soil type parameters, vegetation, aspect etc. In this thesis, we examine the effect of leaf-area-index (LAI) and aspect together on the model simulated actual evapotranspiration (AET) and water balance. Evapotranspiration covers both water evaporation and transpiration. The mesoscale Hydrological Model (mHM) which is fully distributed model process calibration the cycle for five different cases. PET correction driven by neither LAI nor aspect in Case 0. The purpose of adding this case to see the vegetation and aspect effect by comparing other cases. Aspect driven results have been checked in Case 1. Aspect defines direction of slope faces. Aspect is a degree that takes values from 0 to 360. LAI driven results have been checked in Case 2. LAI is the ratio between the vegetation area and the total area. Therefore, LAI is a unitless parameter. LAI, and aspect driven by the weight number results have been checked in Case 3. LAI, and aspect correction numbers are multiplied for PET correction and the results were checked in Case 4. Study area was selected as Main Basin where is part of the Middle Rhine in Germany. Main Basin was selected because gauge station of basin has longterm discharge data without missing values. Static maps (DEM, land use, soil maps etc.), temperature, and precipitation inputs between 2002 and 2014 are added to mHM. Flow values are taken from Global Runoff Data Center (GRDC). Spatial maps are referenced from datasets of moderate resolution imaging spectroradiometer (MODIS), European observation (E-OBS), and The Shuttle Radar Topography Mission (SRTM). To improve water balance and spatial pattern of AET simulations, Kling-Gupta Metric (KGE) and Spatial Efficiency Metric SPAEF are used. For water balance simulation, one outlet gauge is used for calculations with objective function of KGE. For spatial pattern of AET, objective function of SPAEF is used. Best solutions and sensitivity of parameters were selected by sum of objective functions of KGE and SPAEF. In the continuation of study, the Parameter Estimation Tool (PEST) was used for sensitivity analysis and twenty-six parameters were selected. These sensitive parameters selected according to objective functions. Accoording to result, sensitive parameters contain especially LAI, aspect, and soil parameters. Selected parameters calibrated for each case by using Optimization Software Toolkit (OSTRICH) with Pareto Archived Dynamically Dimensioned Search (PADDS) algorithm. According to best result of each case, Case 3, and Case 4 show better performance than other cases both on water balance and spatial pattern of AET. This result show that using both LAI and aspect increase simulation performance of mHM. Case 2 also show better performance than Case 1 and Case 0. In this way, LAI corrected PET made simulation on streamflow and spatial pattern of AET better than aspect corrected case, and not corrected case. Raw PET water balance result is better than aspect corrected one. Best result of Case 3 is closer to the theoretical best sum of SPAEF and KGE than Case 4. That result show that using weight parameter (alphax) exhibited slightly better result than multiplied LAI and aspect corrected case. However, except best solutions Case 4 has better solution than Case 3 for only SPAEF. Additionally, validation with six gauges was done. Case 3 and Case 4 showed better KGE performance than other case. Especially Case 3 similarity with observed discharge is much higher than other cases with six gauges validation. Although the thesis is based on one basin study area, the findings suggest validating and calibrate AET and discharge by using both LAI and aspect.