The effect of climate on the hydrological regime of selected Greek areas with different climate conditions Thesis submitted in partial fulfilment of the requirements of the degree of Doctor rer. nat. of the Faculty of Environment and Natural Resources Albert-Ludwigs-University Freiburg im Breisgau, Germany by Spyridon Paparrizos Freiburg im Breisgau, Germany 2016 Dean: Prof. Dr. Tim Freytag Advisor: Prof. Dr. Andreas Matzarakis 2nd Advisor: Prof. Dr. Markus Weiler 3rd Advisor: Ass. Prof. Dr. Fotios Maris Oral examination date: November 7, 2016 2 To my parents Penelope and Antonios 3 Acknowledgements By finalising the current Thesis, I would like first of all to express my gratitude to my supervisor Professor Dr. Andreas Matzarakis, especially for the continuous and unfailing supervision, as well as his priceless guidance during my life here in Freiburg. The scientific work that we performed together and our talks will be forever depicted in my Thesis, and in my memory. I would also like to thank my second supervisor, Professor Dr. Markus Weiler who has expressed kind interest about my research through our occasional talks. Special thanks need to be given to the Associate Professor Dr. Fotios Maris, my supervisor during my BSc. and MSc. studies and my mentor since my juvenile years in the university. I would be forever grateful for the time he devoted to me, as well as the resources that he provided me and I am sure that we will continue working together. I am thankful to all my colleagues of the Chair of Environmental Meteorology and first of all to PD Dr. Dirk Schindler who gave me the opportunity to work as a research assistant and assist him with some lectures. My fellow colleagues at Hebelstrasse: Dominik Fröhlich, Sven Gebhart, Ronja Vitt, Marcel Gangwisch, Shiqi Yang, Yung-Chang Chen, as well as my former colleagues Dr. Emmanuel Lubango Ndetto, Dr. Christine Ketterer and Dr. Letizia Martinelli for the substantial help that provided to me, our exciting everyday talks, lunches, and occasional feierabends; without you my life in Freiburg would have been less open- minded, less multi-cultural, and probably ennui. I want to thank also Dr. Simeon Potouridis, a colleague from my previous university and a life colleague for his interest, his support and our long talks. He urged me to come and study in Freiburg, and this was one of the best decisions I ever made. He was always there for me and I know that he will continue to be. Although the current Thesis is dedicated to my parents, I feel the need to write a few more words to emphasize their contribution. I want to thank my parents Penelope and Antonios as so my sister Emily for their substantial economic but most important: their moral support. Our parents always believed in my sister and me from the beginning of our lives, they raised us with dignity; they deprived many things in order to provide us everything. I could have never asked for a better family and I am more than thankful for that. I sincerely appreciate the help of all the teachers and professors I had since the early years of my life as each and every one of them contributed their part towards the actualization of the current Thesis, and the achievement of my goals. Finally, I want to say thanks to my friends and everyone I consider as family and have supported me all these years. 4 Summary Assessment of present and future variations of the aspects of the hydrological cycle is essential for the research regarding climate and climate change. The current study is focused on 3 selected agricultural areas widespread in Greece, the Ardas River basin in North-eastern Greece, the Sperchios River basin in central Greece, and the Geropotamos River basin in Crete Island in South Greece that have different climate characteristics due to their location, as well as complex topography. The aim is to analyse the various aspects and procedures of the hydrological cycle, and assess its future variations, responses, and their effect on the hydrological regime of the study areas, with specific focusing on the agricultural productivity. Furthermore, the study aimed to assess the role that topography plays in the formation and spatial distribution of the climate conditions prevailing in the certain study areas. Initially, the study was focused on Potential Evapotranspiration (PET), which is one of the most critical parameters in the research on agro-ecological systems. The computational methods for the estimation of PET vary in data demands from very simple (empirically based), requiring only information based on air temperatures, to complex ones (more physically based) that require data on radiation, air humidity, etc. For this reason, 12 Potential Evapotranspiration formulae were adopted, analysed and inter-compared in terms of their sensitivity regarding their input coefficients. The aim was to compare all the methods and conclude to which empirical PET method(s) better represent the PET results in each area and thus should be adopted and used each time, and which factors influence the results in each case. Subsequently the assessment and comparison of the future variation of annual and seasonal precipitation was performed, as well as the assessment of the future response of streamflow and its impacts on the hydrological regime, in combination with other fundamental aspects of the hydrological cycle. The determination of heat requirements in the first developing phases of plants, that has been expressed as Growing Degree Days (GDD) was implemented, in order to assess the future variation and the spatial distribution of the GDD, and how these can affect the main cultivations in the study areas. Finally, the aridity and drought conditions prevailing in the study areas were estimated. Since a major task and object of the current study was to spatially interpret the various components and procedures, the current study describes a technique for downscaling climatological data in areas with limited data. In cases where the observations from the meteorological network as well as the density are insufficient to cover the entire study area and the researcher is called to operate on a regional or the mesoscale and 5 produce detailed results, the technique can be a helping hand. It constitutes a combination of statistical downscaling through multi-linear regression techniques with the dynamical downscaling through Geographic Information Systems and it was used in order to spatially interpolate with high resolution various climatological procedures in the current study. Future meteorological data were derived and analysed from a number of Regional Climate Models (RCMs) from the ENSEMBLES European project. The climate simulations were performed for the future periods 2021-2050 and 2071-2100, under the A1B and B1 emission scenarios that were developed by the Intergovernmental Panel on Climate Change. Spatial interpolation was performed using the combined dynamical and statistical downscaling technique and the Ordinary Kriging method within ArcGIS 10.2.1. Mann-Kendall statistical control method was used to investigate possible trends, while the Auto-regressive integrated moving average model (ARIMA) was used for forecasting within Mathworks version of 2014a. ArcSWAT ArcGIS extension was used to simulate the future responses of streamflow. Growing Degree Days (GDD) units was adopted to determine the present and future heat requirements of the existing cultivations. Present aridity conditions were estimated using the Aridity index (AI), while the Standardized Precipitation Index (SPI) was used to identify and assess the present and future drought conditions. According to the combined downscaling technique that was proposed in the current study, the results indicated that the current technique delivered very sufficient results as the adjusted coefficient (R2) was appeared with high values in almost every case. Areas characterized by Mediterranean type of climate with hot summers (Csa) showed the strongest presumption against null hypothesis; while areas characterized by a combination of different Mediterranean climate types (Csa and Csb) used the most coefficients in the regression equations and produced relatively good results. Areas facing continental climate conditions also delivered satisfactory results, although most of the examined independent coefficients that were used in the regression analyses were presented with medium presumption against null hypothesis. Summarizing, the described technique can be used for every type of climate in almost every terrain for the accurate representation of various climatological variables in the mesoscale. Regarding the sensitivity analysis of PET formulae, the results indicated that for the areas that face Mediterranean climate conditions, the most appropriate method for the nd estimation of PET was the temperature-based, Hamon's 2 version (PETHam2). Furthermore, the PETHam2 equation was able to estimate PET almost with the same efficiency as the average results of the 12 empirical formulae. For the Ardas River basin, the results indicated that both PETHam2 and PETHam1 can be used to estimate PET satisfactorily. Moreover, the temperature-based equations have proven to produce finer results, followed by the radiation- 6 based equations. On the other hand, PETASCE which is the most common equation, can also be applied occasionally in order to provide satisfactory results. Concerning the results of the integrated analysis and future responses of precipitation, it is expected to be critically decreased for both scenarios, future periods and study areas. Specifically, precipitation is expected to decrease by 32-40% by the end of the century in the Sperchios River basin, by almost 45% in the Ardas River basin, and more than 50% in the Geropotamos River basin. Furthermore, the decrease in precipitation for the Ardas River basin which is characterized by continental climate will be tempered, while in the Sperchios River basin the decrease will be smoother due to the influence of some minor climatic variations in the basins' springs in the highlands where milder conditions occur.