Catchment scaled drought analysis: Integrated climate and hydrological perspectives

The Intergovernmental Panel on Climate Change (IPCC) set up by United Nations Environment Program (UNEP) and World Meteorological Organization (WMO) in 1988 to provide insights for decision and policy-makers with systematic evaluations, impacts, probable risks and options for mitigation and adaptation of climate change. The concerns in increasing of Greenhouse Gases (GHGs) are still alive among scientists e.g. by the April 2018 the concentration of CO2, which itself forms half of the GHGs emissions, passed 410 ppm which explains an increase of 30 percentage comparing to the pre-industrial era. Representative Concentration Pathways (RCPs) are four GHG concentration (not emissions) trajectories adopted by fifth Assessment Report (AR5) of the IPCC in 2014, which overtakes Special Report on Emissions Scenarios (SRES) projections published in 2000. The pathways are developed for climatic modeling and investigations describing four potential climate perspectives, all of them are considered possible depending on how much GHGs are released in the future. The outcomes of Global Climate Models (GCMs) and Regional Climate Models (RCMs) are used for predicting the quantity and variability of different parameters for the future analysis in many fields. Recently, the frequency of extreme climatological events increased dramatically in different parts of the world due to a probable changing climate. According to the IPCC AR5 synthesis report, all RCPs shows an increase in the surface temperature and decrease in precipitation totals for Turkey with the high confidence. The increase in average temperature in accordance with changes in precipitation patterns, decrease in snowy days, increases in heat waves parallel to direct anthropogenic effects like population increase, groundwater over-exploitations and desertification altogether causes the Turkey to face water crises during the 21th century. Therefore, it is one of the hotspots for scientists and policymakers, as it faced with different drought events in many regions during the last decades and enters to the alarming situation in association with climate change. Drought indices are essential for tracking and anticipating drought-related impacts and outcomes. Indices may also play another critical role where they can provide holistic reference for planners or decision-makers. Recently, researchers put efforts into developing new indices applying novel techniques particularly focusing on utilizing the information of two or more indicators (precipitation, temperature, runoff etc.) known as multivariate drought index. The primary objective of this research is to study the drought analysis on a catchment scale by a various approaches over the Konya Closed Basin (KCB). In the first phase of this thesis, development of various well-known drought indices for meteorological, hydrological and agricultural drought types are done in parametric and non-parametric methods. The former proposed meteorological parametric drought indices (SPI and SPEI) are chosen because, firstly they are easy to calculate and secondly, for revisiting the previous studies in the KCB and finally, to show the role of temperature in intensifying drying condition over the study area. Results clarify the role of temperature increase in the region causing to surge the drought severity particularly after 2000. Surprisingly, it is shown that between 2000 and 2010, most of the stations experienced drought in different severities. Further, the nonparametric meteorological, hydrological and agricultural drought indices was utilized to assess the holistic historical drought conditions in the KCB. Moreover, the trends are investigated using Mann-Kendall and Sen's Graphical nonparametric methods on hydrological and agricultural indices. The results indicate various drying trends in different areas on 3- and 6-month timescales. On the other hand, the cross wavelet analysis is assessed to investigate the probable teleconnections of hydrological and agricultural drought indices with large-scale atmospheric circulations, which reveals various significant correlations over the study area. In the next chapter, the historical nonparametric multivariate standardized drought index (NMSDI) is assessed using the information of precipitation and soil moisture via copulas. The bivariate drought characteristics of NMSDI is handled with several copula families to achieve the historical drought return periods on various return period scenarios and drought categories. The copula analysis on historical nonparametric drought index reveals the superiority of BB6 copula family in modelling the joint dependence of drought characteristics over the KCB. Overall, the southeastern and southwestern regions reflect the high-risk area on 3-month time scales as oppose to the north and northwestern parts seems on 6-month time scale on all drought categories in the KCB. To model the joint dependence of NMSDI characteristics via copula for the future period (2019-2100), first the performance of multivariate drought indicators (precipitation and soil moisture series) based on twenty-two EURO-CORDEX ensembles simulations forced by different CMIP5 GCMs are evaluated. To this end, the former univariate metrics using various data statistics are used. In addition the state-of-the-art, entropy-based performance analysis of EURO-CORDEX ensembles simulations are assessed by means of different entropy-based formulations over the KCB. Results indicates the superiority of IPSL-IPSL-CM5A-MR-RCA4, MPI-M-MPI-ESM-LR-REMO2009 and MPI-M-MPI-ESM-LR-RCA4 series in modelling the precipitation and soil moisture series for the future period. The statistical comparison also show that the proposed entropy-based approach could capture the best models among the available series in the KCB very well. Then, to eliminate the probable biases in the selected models, four Bias Correction (BC) methods as parametric, nonparametric and robust nonparametric quantile mapping and, linear scaling method are applied to the chosen series. The performance of each input is evaluated with Root Mean Square Error (RMSE), percentage of RMSE (PRMSE), Mean Absolute Error (MAE), Percentage Bias (PBIAS) and Nash–Sutcliffe Efficiency (NSE) Methods. Results of BC approaches stated the superiority of nonparametric and robust nonparametric BC methods in analysis of selected models over the KCB. The chosen outputs are then used as inputs in two time span, first the period of 2019-2059 and latter one, the period of 2060 to 2100 to generate the NMSDI series distinctively. Moreover, similar to the historical period, the copula analysis is investigated on 3- and 6-month time scales and drought return period based on various drought risk categories are developed. The minimum of the outputs received by drought return period analysis depicts a substantial decrease for the worst categories during the studied periods expressing a considerable drought risks increase over them. Furthermore, stations #242 and #192 illustrates high-risk regions on all drought categories while less risk was predicted for stations located in southern areas based on 3-month NMSDI series derived by IPSL-RCA4/RCP4.5. In order to explore the cause and effect relationship between various drought triggers and drought events, these analyses give a holistic view using different drought indices and statistical analysis. Overall, the results obtained from this study are expected to help achieve an effective drought mitigation strategy for the KCB. A better understanding of the evolution of drought events, integrated characterization of drought, the magnitude of risk it poses, and knowledge of its frequency of occurrence could help humankind to prevent, or at least better prepared for the many devastating consequences droughts can potentially cause.