• Addor, N., , Rössler O. , , Köplin N. , , Huss M. , , Weingartner R. , , and Seibert J. , 2014: Robust changes and sources of uncertainty in the projected hydrological regimes of Swiss catchments. Water Resour. Res., 50, 75417562, doi:10.1002/2014WR015549.

    • Search Google Scholar
    • Export Citation
  • Alpert, P., , Krichak S. , , Shafir H. , , Haim D. , , and Osetinsky I. , 2008: Climatic trends to extremes employing regional modeling and statistical interpretation over the E. Mediterranean. Global Planet. Change, 63, 163170, doi:10.1016/j.gloplacha.2008.03.003.

    • Search Google Scholar
    • Export Citation
  • Beniston, M., , Rebetez M. , , Giorgi F. , , and Marinucci M. R. , 1994: An analysis of regional climate change in Switzerland. Theor. Appl. Climatol., 49, 135159, doi:10.1007/BF00865530.

    • Search Google Scholar
    • Export Citation
  • Black, E., 2009: The impact of climate change on daily precipitation statistics in Jordan and Israel. Atmos. Sci. Lett., 10, 192200, doi:10.1002/asl.233.

    • Search Google Scholar
    • Export Citation
  • Brielmann, H., 2008: Recharge and discharge mechanism and dynamics in the mountainous northern upper Jordan River catchment, Israel. Ph.D. thesis, Faculty of Geosciences, Ludwig Maximilians University Munich, 305 pp. [Available online at https://edoc.ub.uni-muenchen.de/9972/1/Brielmann_Heike.pdf.]

  • Funk, C., and et al. , 2014: A quasi-global precipitation time series for drought monitoring. U.S. Geological Survey Data Series 832, 4 pp., doi:10.3133/ds832.

  • Giorgi, F., 2002: Variability and trends of sub-continental scale surface climate in the twentieth century. Part I: Observations. Climate Dyn., 18, 675691, doi:10.1007/s00382-001-0204-x.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., , Hurrell J. W. , , Marinucci M. R. , , and Beniston M. , 1997: Elevation dependency of the surface climate change signal: A model study. J. Climate, 10, 288296, doi:10.1175/1520-0442(1997)010<0288:EDOTSC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., , Bi X. , , and Pal J. S. , 2004: Mean, interannual variability and trends in a regional climate change experiment over Europe. I. Present-day climate (1961–1990). Climate Dyn., 22, 733756, doi:10.1007/s00382-004-0409-x.

    • Search Google Scholar
    • Export Citation
  • Givati, A., , Lynn B. , , Liu Y. , , and Rimmer A. , 2012: Using the WRF Model in an operational streamflow forecast system for the Jordan River. J. Appl. Meteor. Climatol., 51, 285299, doi:10.1175/JAMC-D-11-082.1.

    • Search Google Scholar
    • Export Citation
  • Goldscheider, N., , and Drew D. , 2007: Methods in Karst Hydrogeology. Taylor and Francis, 278 pp.

  • Green, W., , and Ampt G. , 1911: Studies in soil physics. I. Flow of air and water through soils. J. Agric. Sci., 4, 124, doi:10.1017/S0021859600001441.

    • Search Google Scholar
    • Export Citation
  • Gur, D., , Bar-Matthew M. , , and Sass E. , 2003: Hydrochemistry of the main Jordan River sources: Dan, Banias, and Kezinim springs, north Hula Valley, Israel. Isr. J. Earth Sci., 52, 155178, doi:10.1560/RRMW-9WXD-31VU-MWHN.

    • Search Google Scholar
    • Export Citation
  • Haylock, M., , Hofstra N. , , Klein Tank A. , , Klok E. , , Jones P. , , and New M. , 2008: A European daily high-resolution gridded dataset of surface temperature and precipitation for 1950–2006. J. Geophys. Res., 113, D20119, doi:10.1029/2008JD010201.

    • Search Google Scholar
    • Export Citation
  • Heckl, A., 2011: Impact of climate change on the water availability in the Near East and the upper Jordan River catchment. Ph.D. thesis, University of Augsburg, 181 pp. [Available online at nbn-resolving.de/urn:nbn:de:bvb:384-opus-18095.]

  • Hertig, E., , and Jacobeit J. , 2008: Downscaling future climate change: Temperature scenarios for the Mediterranean area. Global Planet. Change, 63, 127131, doi:10.1016/j.gloplacha.2007.09.003.

    • Search Google Scholar
    • Export Citation
  • Jacob, D., and et al. , 2014: EURO-CORDEX: New high-resolution climate change projections for European impact research. Reg. Environ. Change, 14, 563578, doi:10.1007/s10113-013-0499-2.

    • Search Google Scholar
    • Export Citation
  • Jung, G., , Wagner S. , , and Kunstmann H. , 2012: Joint climate–hydrology modeling: An impact study for the data-sparse environment of the Volta basin in West Africa. Hydrol. Res., 43, 231247, doi:10.2166/nh.2012.044.

    • Search Google Scholar
    • Export Citation
  • Kotlarski, S., and et al. , 2014: Regional climate modeling on European scales: A joint standard evaluation of the EURO-CORDEX RCM ensemble. Geosci. Model Dev., 7, 12971333, doi:10.5194/gmd-7-1297-2014.

    • Search Google Scholar
    • Export Citation
  • Kraller, G., , Warscher M. , , Kunstmann H. , , Vogl S. , , Marke T. , , and Strasser U. , 2012: Water balance estimation in high Alpine terrain by combining distributed modeling and a neural network approach (Berchtesgaden Alps, Germany). Hydrol. Earth Syst. Sci., 16, 19691990, doi:10.5194/hess-16-1969-2012.

    • Search Google Scholar
    • Export Citation
  • Krichak, S. O., , Alpert P. , , Bassat K. , , and Kunin P. , 2007: The surface climatology of the eastern Mediterranean region obtained in a three-member ensemble climate change simulation experiment. Adv. Geosci., 12, 6780, doi:10.5194/adgeo-12-67-2007.

    • Search Google Scholar
    • Export Citation
  • Kunstmann, H., , Heckl A. , , and Rimmer A. , 2006: Physically based distributed hydrological modelling of the upper Jordan catchment and investigation of effective model equations. Adv. Geosci., 9, 123130, doi:10.5194/adgeo-9-123-2006.

    • Search Google Scholar
    • Export Citation
  • Legates, D. R., , and Willmott C. J. , 1990: Mean seasonal and spatial variability in gauge-corrected, global precipitation. Int. J. Climatol., 10, 111127, doi:10.1002/joc.3370100202.

    • Search Google Scholar
    • Export Citation
  • Majone, B., , Bovolo C. I. , , Bellin A. , , Blenkinsop S. , , and Fowler H. J. , 2012: Modeling the impacts of future climate change on water resources for the Gallego River basin (Spain). Water Resour. Res., 48, W01512, doi:10.1029/2011WR010985.

    • Search Google Scholar
    • Export Citation
  • Maraun, D., 2012: Nonstationarities of regional climate model biases in European seasonal mean temperature and precipitation sums. Geophys. Res. Lett., 39, L06706, doi:10.1029/2012GL051210.

    • Search Google Scholar
    • Export Citation
  • Maraun, D., , Osborn T. J. , , and Rust H. W. , 2012: The influence of synoptic airflow on UK daily precipitation extremes. Part II: Regional climate model and E-OBS data validation. Climate Dyn., 39, 287301, doi:10.1007/s00382-011-1176-0.

    • Search Google Scholar
    • Export Citation
  • Mitchell, T. D., , and Jones P. D. , 2005: An improved method of constructing a database of monthly climate observations and associated high-resolution grids. Int. J. Climatol., 25, 693712, doi:10.1002/joc.1181.

    • Search Google Scholar
    • Export Citation
  • Monteith, J., 1975: Vegetation and the Atmosphere. Academic Press, 278 pp.

  • Peleg, N., , Bartov M. , , and Morin E. , 2015: CMIP5-predicted climate shifts over the east Mediterranean: Implications for the transition region between Mediterranean and semi-arid climates. Int. J. Climatol., 35, 21442153, doi:10.1002/joc.4114.

    • Search Google Scholar
    • Export Citation
  • Peschke, G., 1987: Soil moisture and runoff components from a physically founded approach. Acta Hydrophys., 31, 191205.

  • Polade, S. D., , Pierce D. W. , , Cayan D. R. , , Gershunov A. , , and Dettinger M. D. , 2014: The key role of dry days in changing regional climate and precipitation regimes. Sci. Rep., 4, 4364, doi:10.1038/srep04364.

    • Search Google Scholar
    • Export Citation
  • Rimmer, A., , and Salingar Y. , 2006: Modelling precipitation-streamflow processes in karst basin: The case of the Jordan River sources, Israel. J. Hydrol., 331, 524542, doi:10.1016/j.jhydrol.2006.06.003.

    • Search Google Scholar
    • Export Citation
  • Ruti, P., and et al. , 2016: MED-CORDEX initiative for Mediterranean climate studies. Bull. Amer. Meteor. Soc., doi:10.1175/BAMS-D-14-00176.1, in press.

    • Search Google Scholar
    • Export Citation
  • Samuels, R., , Rimmer A. , , and Alpert P. , 2009: Effect of extreme rainfall events on the water resources of the Jordan River, Israel. J. Hydrol., 375, 513523, doi:10.1016/j.jhydrol.2009.07.001.

    • Search Google Scholar
    • Export Citation
  • Samuels, R., , Rimmer A. , , Hartman A. , , Krichak S. , , and Alpert P. , 2010: Climate change impacts on Jordan River flow: Downscaling application from a regional climate model. J. Hydrometeor., 11, 860879, doi:10.1175/2010JHM1177.1.

    • Search Google Scholar
    • Export Citation
  • Samuels, R., , Smiatek G. , , Krichak S. , , Kunstmann H. , , and Alpert P. , 2011: Extreme value indicators in highly resolved climate change simulations for the Jordan River area. J. Geophys. Res., 116, D24123, doi:10.1029/2011JD016322.

    • Search Google Scholar
    • Export Citation
  • Sauter, M., , Geyer T. , , Kovács A. , , and Teutsch G. , 2006: Modellierung der Hydraulik von Karstgrundwasserleitern—Eine Übersicht. Grundwasser, 11, 143156, doi:10.1007/s00767-006-0140-0.

    • Search Google Scholar
    • Export Citation
  • Schulla, J., 2014: Model description WaSiM. Tech. Rep., ETH Zürich, 305 pp. [Available online at http://www.wasim.ch/downloads/doku/wasim/wasim_2015_en.pdf.]

  • Senatore, A., , Mendicino G. , , Smiatek G. , , and Kunstmann H. , 2011: Regional climate change projections and hydrological impact analysis for a Mediterranean basin in southern Italy. J. Hydrol., 399, 7092, doi:10.1016/j.jhydrol.2010.12.035.

    • Search Google Scholar
    • Export Citation
  • Smadi, M., , and Zghoul A. , 2006: A sudden change in rainfall characteristics in Amman, Jordan during the mid 1950s. Amer. J. Environ. Sci., 2, 8491, doi:10.3844/ajessp.2006.84.91.

    • Search Google Scholar
    • Export Citation
  • Smiatek, G., , Kunstmann H. , , and Heckl A. , 2011: High resolution climate change simulations for the Jordan River area. J. Geophys. Res., 116, D16111, doi:10.1029/2010JD015313.

    • Search Google Scholar
    • Export Citation
  • Smiatek, G., , Kunstmann H. , , and Heckl A. , 2014: High-resolution climate change impact analysis on expected future water availability in the upper Jordan catchment and the Middle East. J. Hydrometeor., 15, 15171531, doi:10.1175/JHM-D-13-0153.1.

    • Search Google Scholar
    • Export Citation
  • Somot, S., , Sevault F. , , Déqué M. , , and Crépon M. , 2008: 21st century climate change scenario for the Mediterranean using a coupled atmosphere–ocean regional climate model. Global Planet. Change, 63, 112126, doi:10.1016/j.gloplacha.2007.10.003.

    • Search Google Scholar
    • Export Citation
  • Tramblay, Y., , Ruelland D. , , Somot S. , , Bouaicha R. , , and Servat E. , 2013: High-resolution MED-CORDEX regional climate model simulations for hydrological impact studies: A first evaluation of the ALADIN-Climate model in Morocco. Hydrol. Earth Syst. Sci., 17, 37213739, doi:10.5194/hess-17-3721-2013.

    • Search Google Scholar
    • Export Citation
  • van Genuchten, M. T., 1980: A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Sci. Soc. Amer. J., 44, 892898, doi:10.2136/sssaj1980.03615995004400050002x.

    • Search Google Scholar
    • Export Citation
  • Vano, J. A., , Das T. , , and Lettenmaier D. P. , 2012: Hydrologic sensitivities of Colorado River runoff to changes in precipitation and temperature. J. Hydrometeor., 13, 932949, doi:10.1175/JHM-D-11-069.1.

    • Search Google Scholar
    • Export Citation
  • Vautard, R., and et al. , 2013: The simulation of European heat waves from an ensemble of regional climate models within the EURO-CORDEX project. Climate Dyn., 41, 25552575, doi:10.1007/s00382-013-1714-z.

    • Search Google Scholar
    • Export Citation
  • Wigley, T. M. L., , and Jones P. D. , 1985: Influences of precipitation changes and direct CO2 effects on streamflow. Nature, 314, 149152, doi:10.1038/314149a0.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Investigated subcatchments of the UJR, observational network, and center of the RCM grid cells from the CORDEX simulations. Relief: Shuttle Radar Topography Mission data.

  • View in gallery

    Simulated distribution of the annual mean precipitation in the UJR 1971–2000. MM5_sim denotes WaSiM interpolated precipitation from MM5 run with NCEP–NCAR reanalyses data.

  • View in gallery

    Observed and simulated box plots of the monthly mean rainy season precipitation in the UJR 1971–2000. MM5_sim denotes WaSiM interpolated precipitation from MM5 run with NCEP–NCAR reanalyses data.

  • View in gallery

    Observed and simulated proportions of monthly mean rainy season precipitation 1971–2000. OBS is the mean value of the observed values from UDEL, CRU, E-OBS, and CHIRPS. The bar length shows the bias in the precipitation sum over the considered months related to OBS = 1. OBS and the RCMs apply the entire area depicted in Fig. 1. GCM results apply to a larger area that is different for each GCM. MM5_sim denotes WaSiM interpolated precipitation from MM5 run with NCEP–NCAR reanalyses data.

  • View in gallery

    Distribution of observed and simulated daily runoff 1976–2000.

  • View in gallery

    Observed (OBS) and simulated monthly mean discharge at the Yosef Bridge gauge 1976–2000. OBS* denotes observed values including estimated water removal.

  • View in gallery

    Simulated temperature anomaly 2006–2100 compared to the mean for 1971–2000. The thin red lines show the single-model realizations; the black line shows the multimodel 10-yr running average.

  • View in gallery

    Simulated precipitation anomaly 2006–2100 compared to the mean for 1971–2000. The thin blue lines show the single-model realizations; the black line shows the multimodel 10-yr running average.

  • View in gallery

    Simulated future monthly ensemble mean discharge at the Yosef Bridge gauge. OBS denotes observed and OBS* denotes observed values including estimated water removal, both for 1976–2000.

  • View in gallery

    Observed and simulated distribution of the annual mean Jordan River discharge at the Yosef Bridge gauge. OBS denotes observed values and OBS* denotes observed values including estimated water removal. Simulated values are an ensemble mean of the five RCM models listed in Table 1.

  • View in gallery

    Simulated distribution of daily precipitation over 1 mm in UJR for (a) 1976–2000 and (b) 2071–2100 and simulated distribution of daily discharge at the Yosef Bridge for (c) 1976–2000 and (d) 2071–2100.

  • View in gallery

    Simulated relative runoff change due to relative precipitation changes including evapotranspiration (solid line) and assuming no changes in evapotranspiration [e = 1 in Eq. (2)] (dashed line) for 2071–2100 related to 1976–2000.

  • View in gallery

    Simulated changes in the ETR (a) wrf_sim, (b) cclm_sim, (c) racmo_sim, (d) rca4_sim, and (e) aladin_sim for 2071–2100 related to 1976–2000. Elevation contours over 1400 m, spacing 500 m.

  • View in gallery

    Simulated changes in the total runoff Q (a) wrf_sim, (b) cclm_sim, (c) racmo_sim, (d) rca4_sim, and (e) aladin_sim for 2071–2100 related to 1976–2000. Elevation contours over 1400 m, spacing 500 m.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 62 62 6
PDF Downloads 45 45 3

Expected Future Runoff of the Upper Jordan River Simulated with a CORDEX Climate Data Ensemble

View More View Less
  • 1 Institute of Meteorology and Climate Research, Karlsruhe Institute of Technology, Garmisch-Partenkirchen, Germany
  • | 2 Institute of Meteorology and Climate Research, Karlsruhe Institute of Technology, Garmisch-Partenkirchen, and Institute of Geography, University of Augsburg, Augsburg, Germany
© Get Permissions
Full access

Abstract

Data from five different RCMs run in two experiments from the Coordinated Regional Climate Downscaling Experiment (CORDEX) are applied together with the Water Flow and Balance Simulation Model (WaSiM) to assess the future availability of water in the upper Jordan River. Simulation results for 1976–2000 show that the modeling system was able to reasonably reproduce the observed discharge rates in the partially karstic complex terrain without bias correction of the precipitation input. For the future climate in the area, the applied CORDEX models indicate an increasing annual mean temperature for 2031–60 by 1.8 K above the 1971–2000 mean and by 2.6 K for 2071–2100. The simulated ensemble mean precipitation is predicted to decrease by 16.3% in the first period and 22.1% at the end of the century. In relation to the mean for 1976–2000, the discharge of the upper Jordan River is simulated to decrease by 7.4% until 2060 and by 17.5% until 2100, together with a reduction of high river flow years.

Denotes Open Access content.

Corresponding author address: Gerhard Smiatek, Institute of Meteorology and Climate Research (IMK-IFU), Karlsruhe Institute of Technology, Kreuzeckbahnstr. 19, 82467 Garmisch-Partenkirchen, Germany. E-mail: gerhard.smiatek@kit.edu

Abstract

Data from five different RCMs run in two experiments from the Coordinated Regional Climate Downscaling Experiment (CORDEX) are applied together with the Water Flow and Balance Simulation Model (WaSiM) to assess the future availability of water in the upper Jordan River. Simulation results for 1976–2000 show that the modeling system was able to reasonably reproduce the observed discharge rates in the partially karstic complex terrain without bias correction of the precipitation input. For the future climate in the area, the applied CORDEX models indicate an increasing annual mean temperature for 2031–60 by 1.8 K above the 1971–2000 mean and by 2.6 K for 2071–2100. The simulated ensemble mean precipitation is predicted to decrease by 16.3% in the first period and 22.1% at the end of the century. In relation to the mean for 1976–2000, the discharge of the upper Jordan River is simulated to decrease by 7.4% until 2060 and by 17.5% until 2100, together with a reduction of high river flow years.

Denotes Open Access content.

Corresponding author address: Gerhard Smiatek, Institute of Meteorology and Climate Research (IMK-IFU), Karlsruhe Institute of Technology, Kreuzeckbahnstr. 19, 82467 Garmisch-Partenkirchen, Germany. E-mail: gerhard.smiatek@kit.edu

1. Introduction

The Jordan River, flowing from the Mount Hermon area to the Dead Sea, is the most important freshwater system in the region, and its water is almost fully used by withdrawal from Lake Kinneret, fed to a large extent from the upper Jordan River (UJR) basin. As the countries in the eastern Mediterranean (EM) region show high rates of growth of population and urbanization, the current scarcity of water is expected to worsen in the future, with the additional pressure on the water availability resulting from the expected climate change. Numerous studies addressing the present climate trends have predicted increasing temperatures and decreasing rainfall in large parts of the Mediterranean region (Giorgi 2002; Smadi and Zghoul 2006; Krichak et al. 2007; Senatore et al. 2011; Majone et al. 2012; Tramblay et al. 2013). Regional statistical or dynamical downscaling experiments (e.g., Giorgi et al. 2004; Somot et al. 2008; Alpert et al. 2008; Hertig and Jacobeit 2008; Smiatek et al. 2011; Samuels et al. 2011) indicate future temperature increases in the range of 3–4 K and precipitation decreases of up to 30% at the end of the century. These findings emphasize the need for reliable information on future freshwater availability under the conditions of global climate change. The provision of such information is, however, not simple. Complex terrain (elevation up to 2800 m), ungauged outflow to outside springs in Syria and Lebanon, the lack of meteorological observations, and especially the partially karstic environment make up demanding constraints on any study addressing the future water availability in the UJR area.

Several studies have addressed the hydrology of the UJR. Rimmer and Salingar (2006) developed and applied a conceptual daily precipitation–streamflow model called the Hydrological Model for Karst Environment (HYMKE). Brielmann (2008) investigated the recharge and discharge mechanism and dynamics in the northern UJR catchment. Samuels et al. (2009) applied HYMKE with input from a regional climate model output to assess the impact for 2036–60 of climate change on the Jordan River, predicting a reduction of 10% in precipitation and reduction of 17% in daily mean surface flow. Heckl (2011) coupled the Water Flow and Balance Simulation Model (WaSiM; Schulla 2014) with the Fifth-generation Pennsylvania State University–National Center for Atmospheric Research Mesoscale Model (MM5) driven with ECHAM4 forcings. Recently, Givati et al. (2012) applied HYMKE in an operational streamflow forecast using highly resolved runs of the Weather Research and Forecasting (WRF) Model driven with NCEP–NCAR reanalyses data to derive the meteorology input required by HYMKE.

Smiatek et al. (2014) assessed the future water availability in the UJR region with the distributed hydrology model WaSiM offline coupled with the meteorology model MM5 run at 18.6- and 6.2-km resolutions. This study is hereafter referred to as MM5_sim. The experiment was driven with NCEP–NCAR reanalyses for the years 1971–2000 and with HadCM3 GCM forcings using the A1B scenario for 1971–2099. The present experiment builds on this work and extends it by applying recently available climate change data resulting from two domains of the Coordinated Regional Climate Downscaling Experiment (CORDEX): Europe (EURO-CORDEX) and the Mediterranean (MED-CORDEX). Because MM5_sim utilized only one RCM–GCM combination, it could not address the full range of possible future projections.

The primary scientific aim of the present experiment is to overcome this shortcoming by the application of the hydrological model with meteorology input data from five regional climate models, each driven with boundary forcings from different earth system models (ESMs) applied within phase 5 of the Coupled Model Intercomparison Project (CMIP5). Further scientific questions result from the typical data processing chain used in hydrological climate change impact analysis where selected data output from a complex climate model is passed to subsequent, more specialized hydrological model runs at higher spatial resolution. The specific questions here are 1) can parts of the chain be omitted, for example, by driving the hydrological model with raw CMIP5 data and 2) can the projected runoff change be obtained from the total runoff provided by the CORDEX land surface models? In addition, mechanisms of future precipitation change, uncertainty, ensemble spread, relation between precipitation change and runoff change, and the correction of the precipitation bias present in the input data are discussed.

The article is structured as follows. Section 2 describes the study area, the applied hydrological model, and the simulated meteorology input used. The results obtained for the present and future climate are presented and discussed in section 3, and conclusions are drawn in section 4.

2. Material and methods

a. Investigated catchment

The study area is the upper catchment of the UJR. It extends over 863 km2 within the borderland of Israel, Syria, and Lebanon. The major tributaries (Fig. 1) are the Dan, with a catchment size of 23 km2; the Snir (also called Hasbani), with a catchment size of 600 km2; and the Hermon (also called Banyas), with a catchment size of 158 km2. They are spring fed and originate from the western and southern slopes of Mount Hermon (2814 m MSL). UJR is only a small part of the entire Jordan River catchment, which is 18 300 km2 in size, but it provides the largest contribution to the entire river discharge. The climate in the region is Mediterranean. Summers are hot and dry and winters are mild and wet. While Hermon and Snir respond faster to rainfall, with the highest discharge amounts during the rainy winter season, the Dan does not react directly to rainfall. Its hydrological regime is steadier, with low intra-annual changes. This indicates the existence of and connection to a larger groundwater basin with favorable storage properties in a karstic environment, which is particularly addressed in our modeling approach.

Fig. 1.
Fig. 1.

Investigated subcatchments of the UJR, observational network, and center of the RCM grid cells from the CORDEX simulations. Relief: Shuttle Radar Topography Mission data.

Citation: Journal of Hydrometeorology 17, 3; 10.1175/JHM-D-15-0066.1

b. Meteorological input data

The present study utilizes five meteorology datasets obtained from the EURO-CORDEX and MED-CORDEX downscaling experiments for the emission scenario representative concentration pathway 4.5 (RCP4.5) at a horizontal resolution of 0.11°. Table 1 shows the applied RCMs, the driving boundary CMIP5 ESMs, and the acronyms used to label the hydrology model runs. The center points of the model grid cells are depicted in Fig. 1. Jacob et al. (2014) and Vautard et al. (2013) describe in detail EURO-CORDEX and Ruti et al. (2016) describe MED-CORDEX.

Table 1.

CORDEX climate change data employed in simulations with the hydrology model WaSiM. The rca4_sim is until November 2099.

Table 1.

For the current climate (1971–2000), the investigated models simulate an annual area mean precipitation in the UJR in the range of 891–1222 mm yr−1 (Fig. 2), with an ensemble mean value of 1063 mm yr−1. These values were obtained by interpolating the RCM center point values (see Fig. 1) onto the UJR catchment, applying Thiessen polygons. Because no adequate observational references are available for this area, an exact assessment of the simulation results is difficult. The MM5_sim (MM5 driven with NCEP–NCAR reanalyses) yielded 1044 mm yr−1. Rimmer and Salingar (2006) estimated the climatological annual mean precipitation in the Dan, Snir, and Hermon area to be 958 mm yr−1. The authors extrapolated the available observed precipitation of the southern part of the catchment.

Fig. 2.
Fig. 2.

Simulated distribution of the annual mean precipitation in the UJR 1971–2000. MM5_sim denotes WaSiM interpolated precipitation from MM5 run with NCEP–NCAR reanalyses data.

Citation: Journal of Hydrometeorology 17, 3; 10.1175/JHM-D-15-0066.1

The intermodel range of the simulated annual mean precipitations from the investigated experiments is 330 mm yr−1. The differences in the maximum values for 1971–2000 are larger. The range of the minimum precipitation is from 380 to 705 mm yr−1 (325 mm yr−1) and of the maximum, from 1276 to 1970 mm yr−1 (695 mm yr−1). The ensemble mean interannual spread ranges from 714 to 1257 mm yr−1 and also reveals the large variability of the precipitation in the area.

Reasonable reproduction of the annual mean precipitation values does not sufficiently answer the question if a bias correction of the model projections is needed. Because of the lack of long-term meteorological observations in large areas of the catchment, only gridded datasets interpolated from station observations such as University of Delaware (UDEL; Legates and Willmott 1990), CRU time series, version 3.22 (Mitchell and Jones 2005), at 0.5° resolution, E-OBS (Haylock et al. 2008) at 0.25° resolution, and satellite-measurement-based climatologies can be employed to verify the annual and monthly values and the precipitation seasonality. From the newly available satellite-based climatologies, only the Climate Hazards Group Infrared Precipitation with Station Data (CHIRPS; Funk et al. 2014) sufficiently covers the investigated time period.

Figure 3 shows box plots of the observed and simulated monthly mean precipitation in the rainy season (from November to April) for the period from 1971 to 2000 (CHIRPS from 1981). The considered area is the entire area depicted in Fig. 1. Compared to the observational reference, model simulations reveal higher interquartile ranges and, in most of the cases, also larger total ranges. This can be expected because of the complex terrain and the fact that gridded climatologies contain considerable biases, particular in sparse data regions (Maraun et al. 2012). Median values compare reasonably in December and January and larger disagreement is found in the intermediate fall and spring seasons. The only larger deviation from observations is the comparatively high mean precipitation amount in November found in rca4_sim. January is the month with the highest precipitation amount in three models. In aladin_sim, December and January precipitation amounts are almost equal. In wrf_sim, highest monthly precipitation occurs in February. Maximum biases in the estimated mean precipitation of December–February (DJF) are in the reasonable range from +16% to −18%. Biases in the transitional periods are substantially larger, however, affecting only comparatively small precipitation amounts. These findings justify the omission of bias correction in the present work, which was necessary in MM5_sim and numerous other previous impact studies. In the bias correction of MM5_sim, the quantile–quantile adjustment to precipitation values derived from an MM5 run driven with NCEP–NCAR reanalyses data has been applied. The correction values were estimated for each month.

Fig. 3.
Fig. 3.

Observed and simulated box plots of the monthly mean rainy season precipitation in the UJR 1971–2000. MM5_sim denotes WaSiM interpolated precipitation from MM5 run with NCEP–NCAR reanalyses data.

Citation: Journal of Hydrometeorology 17, 3; 10.1175/JHM-D-15-0066.1

It is here that one additional model available in EURO-CORDEX was not taken into account because, with an area mean of 1800 mm yr−1, it significantly overestimated the annual mean precipitation.

There is no alternative to meteorology data from RCMs in the UJR area. Because of their coarse resolution ranging from 1.4° to over 1.8°, data from the driving models in CMIP5 (Table 1) are not sufficient in an area with the sharp precipitation gradient found in west–east and north–south directions. The simulated annual mean precipitation in the model grid cell covering the UJR ranges from 106 to 650 mm, with an ensemble mean of only 338 mm. Furthermore, the majority of the investigated models fail to reproduce the observed precipitation seasonality (Fig. 4). The models IPSL-CM5A-MR and CNRM-CM5 depict the precipitation maximum in December, EC-EARTH in depicts the maximum in February, and in MPI-ESM-LR precipitation amounts in December and January are almost equal. HadGEM2-ES depicts a correct seasonality, but overestimates the fall precipitation.

Fig. 4.
Fig. 4.

Observed and simulated proportions of monthly mean rainy season precipitation 1971–2000. OBS is the mean value of the observed values from UDEL, CRU, E-OBS, and CHIRPS. The bar length shows the bias in the precipitation sum over the considered months related to OBS = 1. OBS and the RCMs apply the entire area depicted in Fig. 1. GCM results apply to a larger area that is different for each GCM. MM5_sim denotes WaSiM interpolated precipitation from MM5 run with NCEP–NCAR reanalyses data.

Citation: Journal of Hydrometeorology 17, 3; 10.1175/JHM-D-15-0066.1

Simulation results of the applied CORDEX models show much smaller ensemble spread (Fig. 4) than the driving GCM ensemble, indicating a significant role of the regionalization through dynamical downscaling.

c. Hydrological input data

The long-term runoff of the UJR at the outlet of the considered catchment measured at the Yosef Bridge gauge was obtained from the Israeli Hydrological Service (IHS) at daily resolution. As there is substantial water removal prior to the gauging station, the water removals were estimated as in Rimmer and Salingar (2006) by adding the monthly consumptions measured at the pumping stations. No discharge data are available for the outside springs in Lebanon and Syria fed from precipitation in the UJR area. Estimates by Gur et al. (2003) put their total discharge at 250 × 106 m3.

d. The distributed hydrological model

The prime objective of the present study is the quantification of the water balance in the UJR area. The applied hydrological model is WaSiM (Schulla 2014). WaSiM is a deterministic, fully distributed model that applies physically based algorithms for the majority of the process descriptions. These include the infiltration approach after Green and Ampt (1911), the estimation of the saturation time after Peschke (1987), infiltration (by solving the 1D Richards equation), and groundwater flow. In addition, lumped approaches are used in lateral runoff aggregation. Here, the subdivision of the basin into flow time zones is applied to route the surface runoff. Surface runoff retention is done with a single linear storage procedure used in the last flow time zone. The treatment of the translation and retention of the interflow is comparable. Potential evapotranspiration is calculated following the Penman–Monteith (Monteith 1975) method. In calculating the real evapotranspiration, the potential evapotranspiration is reduced, depending on the actual soil moisture content. The applied snow model is based on the temperature–wind index snowmelt method, and the soil moisture dependence of the suction head and the hydraulic conductivity is parameterized as in van Genuchten (1980). A detailed description of the involved parameters and the general setup for the UJR is given by Kunstmann et al. (2006).

The hydrology model has been run at a spatial resolution of 450 m and with a daily time step. The applied input variables were the mean 2-m temperature, precipitation, relative humidity, wind speed, and global radiation. The model calibration involved four parameters in the groundwater submodule, four parameters in the soil submodule, five parameters in the snow submodule, and six parameters accommodating the bypass (see section 2e) and is described in detail in Heckl (2011).

Data from the applied regional climate models were introduced into the hydrological model WaSiM with an interface developed at the Institute of Meteorology and Climate Research–Atmospheric Environmental Research (IMK-IFU) at the Karlsruhe Institute of Technology (KIT) and successfully applied in various studies (e.g., Jung et al. 2012). The interface handles each grid point of the RCM mode as a virtual meteorological station. WaSiM interpolates all input variables from these virtual stations to the 450-m-resolution grid of the hydrological model. As in MM5_sim, the applied interpolation routine was the Thiessen polygon procedure.

e. Adaptation of the hydrological model to UJR

As in MM5_sim, the hydrological model was adapted in order to treat the specific problems of the UJR catchment. Because of the discrepancy between the surface and subsurface catchments, a typical feature of karst systems (Goldscheider and Drew 2007), the water flow in the Snir (600 km2) and Dan (23 km2) catchments cannot be simulated with the conventional WaSiM approach, which is approximated assuming porous media conditions. To cope with the karstic underground and preferential flow paths, extensions to the model based on, for example, the neuronal network approach have recently been applied (Kraller et al. 2012). In the present study (as in MM5_sim), besides the groundwater flow, a bypass within the routing model that moves larger water amounts from the Snir to the Dan catchment was applied. It addresses the duality of the discharge characteristics in karst aquifers: a fast response of highly conductive karst conduits and a delayed drainage of the low-permeability fractured matrix after recharge events (Sauter et al. 2006). The bypass consists of the sum of the direct surface water runoff and the interflow of the Snir. It is directly passed to the Dan subcatchment in a fictitious channel. The channel parameters have been calibrated ensuring a reasonable reproduction of the Dan spring discharge characteristics. A detailed description is given in Heckl (2011).

f. Hydrology model runs

The hydrological model has been run with the meteorology input from the five applied RCMs for the period from 1 January 1971 to 31 December 2010 (rca4_sim until 30 November 2099). To assure direct comparability to MM5_sim, the investigation period for the present climate is from 1 January 1976 to 31 December 2000, allowing for a 5-yr spinup time. The considered future periods are 2031–60 and 2071–2100. Because RCA4 provides data only until 30 November 2099, the mean values for December and the annual mean values derived from rca4_sim include in the second investigated period only 29 values. In all runs, approximately 30% of the routed water was abstracted from the system, accounting for water outflow to ungauged springs in Lebanon and Syria. This is the mean value applied in the simulation MM5_sim, where approximate abstractions of 0.26, 0.3, and 0.33 were applied.

3. Results

a. Present-day climate

The observed annual mean discharge of the UJR at the Yosef Bridge gauge for 1976–2000 is 394.8 × 106 m3 and with the estimated water removal prior to the gauging station, 481.5 × 106 m3. Table 2 shows the statistical analysis of the annual streamflow. Even without a bias correction for the simulated precipitation, the applied models reproduce the mean discharge with a range from 337.2 × 106 (cclm_sim) to 581.2 × 106 m3 (racmo_sim). The ensemble mean runoff value is 470.3 × 106 m3 with an overestimation of the minimum and underestimation of the maximum flow. This is within the range of the results obtained by Rimmer and Salingar (2006) with HYMKE. The authors estimate the mean annual discharge to be 502 × 106 m3, arguing, however, that 109 × 106 m3 are probably contributed to outside springs in Syria. The mean runoff value in MM5_sim was 491.9 × 106 m3.

Table 2.

Observed (OBS) and simulated present and future annual discharge statistics at the Yosef Bridge gauge (× 106 m3). OBS* denotes observed values including estimated water removal. ENS is the ensemble mean. Mean values of MM5_sim are added for comparison.

Table 2.

Figure 5 illustrates the distributions of the observed and simulated daily runoff of the UJR at the Yosef bridge gauge for 1976–2000. The observed mean distribution of the daily discharge is reproduced with coefficients of determination R2 of 0.75 for aladin_sim, 0.49 for cclm_sim, 0.51 for racmo_sim, 0.74 for rca4_sim, and 0.87 for wrf_sim. The observed runoff here was assumed to be the average value of the observed and estimated runoffs, including water removals. The R2 value of the ensemble mean is 0.79. With the exception of cclm_sim and aladin_sim, all the others overestimate the observed daily mean by up to 30%. The quartile values are within a reasonable range. A larger variation can be seen in the extreme daily discharges.

Fig. 5.
Fig. 5.

Distribution of observed and simulated daily runoff 1976–2000.

Citation: Journal of Hydrometeorology 17, 3; 10.1175/JHM-D-15-0066.1

The monthly courses of the observed and simulated runoffs of the UJR for 1976–2000 are shown in Fig. 6. Taking into account the inherent constraints of the discharge simulation in the UJR area, such as the complex terrain, the karstic environment, and the ungauged outflow, the applied hydrological model reasonably reproduces the observed runoff with all applied RCM input data. This is especially true for the model ensemble mean value, as depicted in Fig. 9 (described in greater detail below). Thus, it builds a solid base for the analysis of the future availability of water in the UJR area.

Fig. 6.
Fig. 6.

Observed (OBS) and simulated monthly mean discharge at the Yosef Bridge gauge 1976–2000. OBS* denotes observed values including estimated water removal.

Citation: Journal of Hydrometeorology 17, 3; 10.1175/JHM-D-15-0066.1

The application of a calibrated hydrological model run with more adequate topography, soil, and land-use data, and the consideration of specific processes present in the area allows for a detailed view of the catchment and its subbasins. This is not possible referring to runoff generated by the land surface models applied in the CORDEX RCMs. Nevertheless, a question arises: are the results significantly different? Because of the complex terrain in the area, the coarse model resolution and the fact that an exact delineation of the considered catchment is not possible, CORDEX RCMs (see Table 1) show substantially lower runoff values. For example, for 1976–2000 ALADIN 5.2 reveals 245 × 106 m3, RACMO22E reveals 336 × 106 m3, and RCA4 reveals 320 × 106 m3, which is up to 50% less than the runoff generated by the corresponding WaSiM. WaSiM runoff is not only higher but also close to observations. Thus, the application of a hydrological model clearly increases the reliability of the water availability study.

b. Future climate

In the analysis of the expected future climate, we consider all RCM grid cells depicted in Fig. 1. Figure 7 shows temperature anomalies from the annual mean during 1971–2000 for the UJR area and the 10-yr running average of the model ensemble mean in which the single models were equally weighted. It reveals a simulated increase of the annual mean temperature in the UJR area in the range from 2 to over 4 K at the end of the century. The ensemble mean values increase considerably until 2070. Only a small increase can be observed after 2070. However, the intermodel variability increases in this period, indicating a larger uncertainty. The corresponding precipitation anomalies are depicted in Fig. 8. The figure shows substantial interannual variability present in the area. The simulated ensemble precipitation decreases from 2030 continually until the end of the century, when it reaches about 140 mm yr−1. This is approximately 20% less than the annual mean precipitation of the period from 1971 to 2000. This is within the range of the previous studies: based on simulations of RegCM3 driven with boundary forcings from ECHAM5 (A1B scenario), Samuels et al. (2010) estimated the precipitation reduction for 2036–60 to be 10% from the historical period, 1980–2004. In MM5_sim, the simulated precipitation reduction in the UJR area was in the range between 10% and 35%. Peleg et al. (2015) analyzed the effects of climate change on the EM region based on four CMIP5 models and concluded that by the middle of the century there would be a warming by 1.1–2.6 K and a 10%–22% reduction in wet events.

Fig. 7.
Fig. 7.

Simulated temperature anomaly 2006–2100 compared to the mean for 1971–2000. The thin red lines show the single-model realizations; the black line shows the multimodel 10-yr running average.

Citation: Journal of Hydrometeorology 17, 3; 10.1175/JHM-D-15-0066.1

Fig. 8.
Fig. 8.

Simulated precipitation anomaly 2006–2100 compared to the mean for 1971–2000. The thin blue lines show the single-model realizations; the black line shows the multimodel 10-yr running average.

Citation: Journal of Hydrometeorology 17, 3; 10.1175/JHM-D-15-0066.1

The simulated changes in the seasonal mean temperature and annual mean precipitation for 2031–60 and 2071–2100 compared to 1971–2000 are shown in Table 3. The applied models agree on a temperature increase with ranges in the winter season between 1.4–2.2 K in the first and 1.5–3.0 K in the second period. They disagree on the change in the amount of seasonal precipitation for 2031–60: some models simulate increases of up to 9.6%. For the second investigated period, from 2071 to 2100, all models indicate precipitation reduction in all seasons. In the wet winter season (DJF) with the largest amount of precipitation, the simulated reductions range from 7.5% to 44%. In the spring [March–May (MAM)], the range is from 3.5% to 28.3%. The simulated ensemble annual mean temperature increase is 1.8 K for 2031–60 and 2.6 K for 2071–2100, both related to the 1971–2000 mean value.

Table 3.

Simulated change in the seasonal [DJF, MAM, June–August (JJA), and September–November (SON)] mean temperature and annual mean precipitation compared to 1971–2000. Average over grid points shown in Fig. 1.

Table 3.

In the UJR catchment itself, the annual mean precipitation actually seen by the hydrology model, that is, after interpolation by applying Thiessen polygons, is simulated to decrease by 16.3% for the first investigated future period and by 22.1% for the second. Single-model realizations reveal in the respective periods precipitation decreases in the range from 5.0% to 23.7% and from 10.6% to 35.5%.

Black (2009) associated future precipitation changes in Jordan and Israel with the reduction in the strength of the Mediterranean storm track. From an investigation of 28 CMIP5 models, Polade et al. (2014) conclude that in the Mediterranean region a combination of an increase of dry days and changes in precipitation intensity on wet days drives the changes in annual precipitation. The present investigation supports these findings. Compared to 1971–2000, all models simulate an increase of the number of dry days up to 10 days in 2031–60 and up to 15 days at the end of the century.

The simulated ensemble mean discharge values at the Yosef Bridge gauge are shown in Figs. 9 and 10. Statistics of the single-model realizations are presented in Table 2. With exception of cclm_sim, which shows an increase of 2.5%, the simulations indicate a decrease in the annual mean discharge ranging from 4.4% to 17.2% for 2031–60. The ensemble mean decrease is here 7.4%. All values are related to the mean of the period 1976–2000. For 2071–2100, the discharge decrease intensifies to an ensemble mean of 17.5%, with a range from +1.9% to a decrease of 27.7%.

Fig. 9.
Fig. 9.

Simulated future monthly ensemble mean discharge at the Yosef Bridge gauge. OBS denotes observed and OBS* denotes observed values including estimated water removal, both for 1976–2000.

Citation: Journal of Hydrometeorology 17, 3; 10.1175/JHM-D-15-0066.1

Fig. 10.
Fig. 10.

Observed and simulated distribution of the annual mean Jordan River discharge at the Yosef Bridge gauge. OBS denotes observed values and OBS* denotes observed values including estimated water removal. Simulated values are an ensemble mean of the five RCM models listed in Table 1.

Citation: Journal of Hydrometeorology 17, 3; 10.1175/JHM-D-15-0066.1

Compared to the observed runoff, 481.5 × 106 m3 with estimated water removals, this means a reduction of the runoff of 34.6 × 106 (7%) for 2031–60 and 82.2 × 106 m3 (17%) for 2071–2100. Those changes are associated with a reduction of the maximum annual discharges (Fig. 10).

Figure 11 shows the distributions of daily precipitation intensity over 1 mm and distributions of daily discharge values for 1976–2000 and 2071–2100. All models depict decreases in both days with very high precipitation and days with high discharges. This is contrary to the findings of Samuels et al. (2011), who expect increases in the amount of rain at extreme events in the future, however, considering a larger area than the UJR. The simulated ensemble mean reduction of maximum daily precipitation is −13%, with intermodel range from −2% to −24%. The corresponding reduction of maximum discharge is −25% for the ensemble mean and intermodel range is from 0% to −39%.

Fig. 11.
Fig. 11.

Simulated distribution of daily precipitation over 1 mm in UJR for (a) 1976–2000 and (b) 2071–2100 and simulated distribution of daily discharge at the Yosef Bridge for (c) 1976–2000 and (d) 2071–2100.

Citation: Journal of Hydrometeorology 17, 3; 10.1175/JHM-D-15-0066.1

The results obtained in the present study are in line with MM5_sim, which indicated a discharge decrease by 12% until 2060 and by 26% until 2099. Samuels et al. (2010) show in their experiments with HYMKE driven with RCM data from RegCM3 a reduction for 2036–80 in the streamflow by 17.2% compared to the 1980–2005 mean value. The authors conclude in a nonlinear response of the system to changes in precipitation.

Considering the entire simulation period from 1976 to 2100 in all simulations, the water balance
e1
where P is the precipitation, Q is the runoff, E is the evapotranspiration, and ΔS is the storage change, is closed with a negative change in the water storage, ranging from −3.8% (knmi_sim) to −9.7% (cclm_sim), both related to the simulated total precipitation. Also, there is a comparatively large spread, from 25% to 36%, in the evapotranspiration proportion. Rimmer and Salingar (2006) estimated the evapotranspiration in UJR for the present climate to be 23.5%, and Samuels et al. (2009) estimated it to be approximately 19%.
To further address the role of future changes in precipitation and evapotranspiration, we follow Wigley and Jones (1985), who presented a simple model to assess the relative sensitivity of runoff changes r to changes in precipitation and evapotranspiration as a function of runoff ratio , relative change in precipitation , and relative change in evapotranspiration :
e2
where subscripts 0 and 1 denotes present and future values, respectively, and is assumed to remain constant. Single-model realizations differ in from 0.52 (cclm_sim) to 0.78 (racmo_sim) and r from 0.78 to 1.03 for 2031–60 and from 0.66 to 0.97 in 2071–2100.

Figure 12 shows the simulated relative changes in r due to changes in precipitation ratio. To reveal the role of evapotranspiration, the figure also depicts changes in r assuming no changes in evapotranspiration [e = 1 in Eq. (2)]. With runoff ratio getting smaller, slopes of the lines increase, indicating larger effects of precipitation changes on runoff. In general, this effect is moderate in all models, for example, in comparison to arid areas with a small runoff ratio where precipitation changes have an amplified effect in runoff. The cclm_sim shows the highest sensitivity to precipitation changes (dashed line); relative change in runoff is, however, close to 1 for 2071–2100. The reason is a significant decrease in future evapotranspiration. Other models show relative changes in runoff between 0.66 and 0.83 in 2071–2100. In wrf_sim and rca4_sim those changes are related to decreases in both precipitation and evapotranspiration, and in racmo_sim and aladin_sim they are related to decreases in precipitation only. Thus, as expected, precipitation change plays a key role in future discharge of the UJR. On the other hand the models show large differences in the present and simulated future evapotranspiration changes (see Fig. 13), and therefore, further research on this issue in the complex UJR region is required.

Fig. 12.
Fig. 12.

Simulated relative runoff change due to relative precipitation changes including evapotranspiration (solid line) and assuming no changes in evapotranspiration [e = 1 in Eq. (2)] (dashed line) for 2071–2100 related to 1976–2000.

Citation: Journal of Hydrometeorology 17, 3; 10.1175/JHM-D-15-0066.1

Fig. 13.
Fig. 13.

Simulated changes in the ETR (a) wrf_sim, (b) cclm_sim, (c) racmo_sim, (d) rca4_sim, and (e) aladin_sim for 2071–2100 related to 1976–2000. Elevation contours over 1400 m, spacing 500 m.

Citation: Journal of Hydrometeorology 17, 3; 10.1175/JHM-D-15-0066.1

Figure 14 shows the spatial distribution of the simulated change in the mean total discharge 2071–2100 to 1976–2000. The models agree on the largest reduction in the northern part of the catchment and in the Mount Hermon area. This can be related to the fact that lower elevations might even see precipitation increases in the future, while higher elevations depict a much higher precipitation decrease. Smiatek et al. (2011) found a clear elevation dependency of the climate change signal in EM. Higher elevations are simulated to experience a higher temperature increase and higher precipitation decrease than lower elevations. Such elevation dependency is also reported for the European Alps (Giorgi et al. 1997; Beniston et al. 1994).

Fig. 14.
Fig. 14.

Simulated changes in the total runoff Q (a) wrf_sim, (b) cclm_sim, (c) racmo_sim, (d) rca4_sim, and (e) aladin_sim for 2071–2100 related to 1976–2000. Elevation contours over 1400 m, spacing 500 m.

Citation: Journal of Hydrometeorology 17, 3; 10.1175/JHM-D-15-0066.1

The findings of this study are influenced by uncertainty resulting from modeling deficiencies and future emissions projections. Kotlarski et al. (2014) evaluated the EURO-CORDEX models driven with ERA-Interim data stating typical seasonally and regionally averaged biases of ±40% for precipitation and 1.5°C for temperature. In the EM region, all investigated models revealed a cold bias and an underestimation of the precipitation of the winter season larger than 40%.

Another contribution to the uncertainty results from the model sensitivity to changes in precipitation and temperature. Vano et al. (2012) investigated variations in runoff with respect to precipitation (elasticity) and temperature (sensitivity). They defined precipitation elasticity ϵ as
e3
where is the annual mean runoff obtained with multiplicative perturbations of precipitation P, and the temperature sensitivity s as
e4
where is the annual mean runoff obtained with additive perturbations of temperature T. The perturbations in P were 70%, 80%, 90%, 100%, and 110% with Δ = 1% and in T were 0°, 1°, 2°, and 3° with Δ = 0.1°C. WaSiM runs with perturbed P and T were performed with the wrf_sim data input for the period from 1976 to 2005. The results reveal ϵ values between 1.3 and 1.9 and temperature sensitivity of 5% (°C)−1. In their experiments with five hydrological models Vano et al. (2012), report ϵ values from 2 to 6 and temperature sensitivities from −2.8% to −8.4% (°C)−1. WaSiM shows comparable elasticity and sensitivity to changes in precipitation and temperature.

One of the largest societal challenges in the context of climate change is the delineation of adaptation strategies. Especially difficult to handle is the problem of how uncertainties influence the decision process. One of the solutions to this problem is the quantification of the future changes across a large matrix of factors of influence. Addor et al. (2014) investigated discharge projections resulting from the factorial combination of three emission scenarios, 10–20 regional climate models, two postprocessing methods, and three hydrological models of different complexity in six different Swiss catchments demonstrating robust regime changes that emerge despite the projection uncertainty. With results based on four GCMs and five RCMs, the present work contributes to such a matrix that can help in decision-making under uncertainty.

4. Conclusions

Regional climate–hydrology simulations were performed with the physically based distributed hydrological model WaSiM for the upper Jordan River basin with meteorology input from the five dynamical downscaling experiments EURO-CORDEX and MED-CORDEX. The RCM used five different forcings from CMIP5 ESMs, applying the RCP4.5 scenario. The RCM data were used without a bias correction, which indicates a significant improvement compared to the previous experiments. It avoids inconsistencies in the energy fluxes, number of precipitation events, and the required stationarity of the RCM bias that was questioned in some recent studies (Maraun 2012; Tramblay et al. 2013).

As in MM5_sim, the hydrological model WaSiM was successfully run for the UJR, reasonably simulating the water balance in the demanding catchment area with an artificial bypass approach that allows simulation of the karst aquifer and the Dan spring as well as with a water abstraction concept, introduced to take into account water flowing to ungauged outside springs.

The results obtained indicate substantial changes in the future climate of the UJR area. The applied CORDEX models reveal increasing annual mean temperatures, 1.8 K above the 1971–2000 mean for 2031–60 and 2.6 K higher for 2071–2100. The simulated ensemble mean precipitation is simulated to reduce by 16.7% in the first period and 22.1% at the end of the century. Related to the 1976–2000 mean value, the discharge of the UJR is simulated to decrease by 7.4% for 2031–60 and by 17.5% for 2071–2100.

The results of the five climate–hydrology simulations support the conclusions drawn in previous studies and thus imply greater confidence in future projections that the future water resources management in the region is likely to receive additional pressure from the reduced water availability.

Acknowledgments

We acknowledge the World Climate Research Programme’s Working Group on Regional Climate, and the Working Group on Coupled Modelling, former coordinating body of CORDEX and responsible panel for CMIP5. We also thank the climate modeling groups (listed in Table 1 of this paper) for producing and making available their model output. We also acknowledge the Earth System Grid Federation infrastructure, an international effort led by the U.S. Department of Energy’s Program for Climate Model Diagnosis and Intercomparison, the European Network for Earth System Modelling, and other partners in the Global Organisation for Earth System Science Portals (GO-ESSP). The simulations used in this work were downloaded from the MED-CORDEX database (www.medcordex.eu) and Earth System Grid Federation (ESGF; http://www.euro-cordex.net). The authors thank A. Rimmer (Kinneret Limnological Laboratory) and A. Givati (Israeli Hydrological Service) for sharing expert knowledge on the hydrology of the UJR.

REFERENCES

  • Addor, N., , Rössler O. , , Köplin N. , , Huss M. , , Weingartner R. , , and Seibert J. , 2014: Robust changes and sources of uncertainty in the projected hydrological regimes of Swiss catchments. Water Resour. Res., 50, 75417562, doi:10.1002/2014WR015549.

    • Search Google Scholar
    • Export Citation
  • Alpert, P., , Krichak S. , , Shafir H. , , Haim D. , , and Osetinsky I. , 2008: Climatic trends to extremes employing regional modeling and statistical interpretation over the E. Mediterranean. Global Planet. Change, 63, 163170, doi:10.1016/j.gloplacha.2008.03.003.

    • Search Google Scholar
    • Export Citation
  • Beniston, M., , Rebetez M. , , Giorgi F. , , and Marinucci M. R. , 1994: An analysis of regional climate change in Switzerland. Theor. Appl. Climatol., 49, 135159, doi:10.1007/BF00865530.

    • Search Google Scholar
    • Export Citation
  • Black, E., 2009: The impact of climate change on daily precipitation statistics in Jordan and Israel. Atmos. Sci. Lett., 10, 192200, doi:10.1002/asl.233.

    • Search Google Scholar
    • Export Citation
  • Brielmann, H., 2008: Recharge and discharge mechanism and dynamics in the mountainous northern upper Jordan River catchment, Israel. Ph.D. thesis, Faculty of Geosciences, Ludwig Maximilians University Munich, 305 pp. [Available online at https://edoc.ub.uni-muenchen.de/9972/1/Brielmann_Heike.pdf.]

  • Funk, C., and et al. , 2014: A quasi-global precipitation time series for drought monitoring. U.S. Geological Survey Data Series 832, 4 pp., doi:10.3133/ds832.

  • Giorgi, F., 2002: Variability and trends of sub-continental scale surface climate in the twentieth century. Part I: Observations. Climate Dyn., 18, 675691, doi:10.1007/s00382-001-0204-x.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., , Hurrell J. W. , , Marinucci M. R. , , and Beniston M. , 1997: Elevation dependency of the surface climate change signal: A model study. J. Climate, 10, 288296, doi:10.1175/1520-0442(1997)010<0288:EDOTSC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., , Bi X. , , and Pal J. S. , 2004: Mean, interannual variability and trends in a regional climate change experiment over Europe. I. Present-day climate (1961–1990). Climate Dyn., 22, 733756, doi:10.1007/s00382-004-0409-x.

    • Search Google Scholar
    • Export Citation
  • Givati, A., , Lynn B. , , Liu Y. , , and Rimmer A. , 2012: Using the WRF Model in an operational streamflow forecast system for the Jordan River. J. Appl. Meteor. Climatol., 51, 285299, doi:10.1175/JAMC-D-11-082.1.

    • Search Google Scholar
    • Export Citation
  • Goldscheider, N., , and Drew D. , 2007: Methods in Karst Hydrogeology. Taylor and Francis, 278 pp.

  • Green, W., , and Ampt G. , 1911: Studies in soil physics. I. Flow of air and water through soils. J. Agric. Sci., 4, 124, doi:10.1017/S0021859600001441.

    • Search Google Scholar
    • Export Citation
  • Gur, D., , Bar-Matthew M. , , and Sass E. , 2003: Hydrochemistry of the main Jordan River sources: Dan, Banias, and Kezinim springs, north Hula Valley, Israel. Isr. J. Earth Sci., 52, 155178, doi:10.1560/RRMW-9WXD-31VU-MWHN.

    • Search Google Scholar
    • Export Citation
  • Haylock, M., , Hofstra N. , , Klein Tank A. , , Klok E. , , Jones P. , , and New M. , 2008: A European daily high-resolution gridded dataset of surface temperature and precipitation for 1950–2006. J. Geophys. Res., 113, D20119, doi:10.1029/2008JD010201.

    • Search Google Scholar
    • Export Citation
  • Heckl, A., 2011: Impact of climate change on the water availability in the Near East and the upper Jordan River catchment. Ph.D. thesis, University of Augsburg, 181 pp. [Available online at nbn-resolving.de/urn:nbn:de:bvb:384-opus-18095.]

  • Hertig, E., , and Jacobeit J. , 2008: Downscaling future climate change: Temperature scenarios for the Mediterranean area. Global Planet. Change, 63, 127131, doi:10.1016/j.gloplacha.2007.09.003.

    • Search Google Scholar
    • Export Citation
  • Jacob, D., and et al. , 2014: EURO-CORDEX: New high-resolution climate change projections for European impact research. Reg. Environ. Change, 14, 563578, doi:10.1007/s10113-013-0499-2.

    • Search Google Scholar
    • Export Citation
  • Jung, G., , Wagner S. , , and Kunstmann H. , 2012: Joint climate–hydrology modeling: An impact study for the data-sparse environment of the Volta basin in West Africa. Hydrol. Res., 43, 231247, doi:10.2166/nh.2012.044.

    • Search Google Scholar
    • Export Citation
  • Kotlarski, S., and et al. , 2014: Regional climate modeling on European scales: A joint standard evaluation of the EURO-CORDEX RCM ensemble. Geosci. Model Dev., 7, 12971333, doi:10.5194/gmd-7-1297-2014.

    • Search Google Scholar
    • Export Citation
  • Kraller, G., , Warscher M. , , Kunstmann H. , , Vogl S. , , Marke T. , , and Strasser U. , 2012: Water balance estimation in high Alpine terrain by combining distributed modeling and a neural network approach (Berchtesgaden Alps, Germany). Hydrol. Earth Syst. Sci., 16, 19691990, doi:10.5194/hess-16-1969-2012.

    • Search Google Scholar
    • Export Citation
  • Krichak, S. O., , Alpert P. , , Bassat K. , , and Kunin P. , 2007: The surface climatology of the eastern Mediterranean region obtained in a three-member ensemble climate change simulation experiment. Adv. Geosci., 12, 6780, doi:10.5194/adgeo-12-67-2007.

    • Search Google Scholar
    • Export Citation
  • Kunstmann, H., , Heckl A. , , and Rimmer A. , 2006: Physically based distributed hydrological modelling of the upper Jordan catchment and investigation of effective model equations. Adv. Geosci., 9, 123130, doi:10.5194/adgeo-9-123-2006.

    • Search Google Scholar
    • Export Citation
  • Legates, D. R., , and Willmott C. J. , 1990: Mean seasonal and spatial variability in gauge-corrected, global precipitation. Int. J. Climatol., 10, 111127, doi:10.1002/joc.3370100202.

    • Search Google Scholar
    • Export Citation
  • Majone, B., , Bovolo C. I. , , Bellin A. , , Blenkinsop S. , , and Fowler H. J. , 2012: Modeling the impacts of future climate change on water resources for the Gallego River basin (Spain). Water Resour. Res., 48, W01512, doi:10.1029/2011WR010985.

    • Search Google Scholar
    • Export Citation
  • Maraun, D., 2012: Nonstationarities of regional climate model biases in European seasonal mean temperature and precipitation sums. Geophys. Res. Lett., 39, L06706, doi:10.1029/2012GL051210.

    • Search Google Scholar
    • Export Citation
  • Maraun, D., , Osborn T. J. , , and Rust H. W. , 2012: The influence of synoptic airflow on UK daily precipitation extremes. Part II: Regional climate model and E-OBS data validation. Climate Dyn., 39, 287301, doi:10.1007/s00382-011-1176-0.

    • Search Google Scholar
    • Export Citation
  • Mitchell, T. D., , and Jones P. D. , 2005: An improved method of constructing a database of monthly climate observations and associated high-resolution grids. Int. J. Climatol., 25, 693712, doi:10.1002/joc.1181.

    • Search Google Scholar
    • Export Citation
  • Monteith, J., 1975: Vegetation and the Atmosphere. Academic Press, 278 pp.

  • Peleg, N., , Bartov M. , , and Morin E. , 2015: CMIP5-predicted climate shifts over the east Mediterranean: Implications for the transition region between Mediterranean and semi-arid climates. Int. J. Climatol., 35, 21442153, doi:10.1002/joc.4114.

    • Search Google Scholar
    • Export Citation
  • Peschke, G., 1987: Soil moisture and runoff components from a physically founded approach. Acta Hydrophys., 31, 191205.

  • Polade, S. D., , Pierce D. W. , , Cayan D. R. , , Gershunov A. , , and Dettinger M. D. , 2014: The key role of dry days in changing regional climate and precipitation regimes. Sci. Rep., 4, 4364, doi:10.1038/srep04364.

    • Search Google Scholar
    • Export Citation
  • Rimmer, A., , and Salingar Y. , 2006: Modelling precipitation-streamflow processes in karst basin: The case of the Jordan River sources, Israel. J. Hydrol., 331, 524542, doi:10.1016/j.jhydrol.2006.06.003.

    • Search Google Scholar
    • Export Citation
  • Ruti, P., and et al. , 2016: MED-CORDEX initiative for Mediterranean climate studies. Bull. Amer. Meteor. Soc., doi:10.1175/BAMS-D-14-00176.1, in press.

    • Search Google Scholar
    • Export Citation
  • Samuels, R., , Rimmer A. , , and Alpert P. , 2009: Effect of extreme rainfall events on the water resources of the Jordan River, Israel. J. Hydrol., 375, 513523, doi:10.1016/j.jhydrol.2009.07.001.

    • Search Google Scholar
    • Export Citation
  • Samuels, R., , Rimmer A. , , Hartman A. , , Krichak S. , , and Alpert P. , 2010: Climate change impacts on Jordan River flow: Downscaling application from a regional climate model. J. Hydrometeor., 11, 860879, doi:10.1175/2010JHM1177.1.

    • Search Google Scholar
    • Export Citation
  • Samuels, R., , Smiatek G. , , Krichak S. , , Kunstmann H. , , and Alpert P. , 2011: Extreme value indicators in highly resolved climate change simulations for the Jordan River area. J. Geophys. Res., 116, D24123, doi:10.1029/2011JD016322.

    • Search Google Scholar
    • Export Citation
  • Sauter, M., , Geyer T. , , Kovács A. , , and Teutsch G. , 2006: Modellierung der Hydraulik von Karstgrundwasserleitern—Eine Übersicht. Grundwasser, 11, 143156, doi:10.1007/s00767-006-0140-0.

    • Search Google Scholar
    • Export Citation
  • Schulla, J., 2014: Model description WaSiM. Tech. Rep., ETH Zürich, 305 pp. [Available online at http://www.wasim.ch/downloads/doku/wasim/wasim_2015_en.pdf.]

  • Senatore, A., , Mendicino G. , , Smiatek G. , , and Kunstmann H. , 2011: Regional climate change projections and hydrological impact analysis for a Mediterranean basin in southern Italy. J. Hydrol., 399, 7092, doi:10.1016/j.jhydrol.2010.12.035.

    • Search Google Scholar
    • Export Citation
  • Smadi, M., , and Zghoul A. , 2006: A sudden change in rainfall characteristics in Amman, Jordan during the mid 1950s. Amer. J. Environ. Sci., 2, 8491, doi:10.3844/ajessp.2006.84.91.

    • Search Google Scholar
    • Export Citation
  • Smiatek, G., , Kunstmann H. , , and Heckl A. , 2011: High resolution climate change simulations for the Jordan River area. J. Geophys. Res., 116, D16111, doi:10.1029/2010JD015313.

    • Search Google Scholar
    • Export Citation
  • Smiatek, G., , Kunstmann H. , , and Heckl A. , 2014: High-resolution climate change impact analysis on expected future water availability in the upper Jordan catchment and the Middle East. J. Hydrometeor., 15, 15171531, doi:10.1175/JHM-D-13-0153.1.

    • Search Google Scholar
    • Export Citation
  • Somot, S., , Sevault F. , , Déqué M. , , and Crépon M. , 2008: 21st century climate change scenario for the Mediterranean using a coupled atmosphere–ocean regional climate model. Global Planet. Change, 63, 112126, doi:10.1016/j.gloplacha.2007.10.003.

    • Search Google Scholar
    • Export Citation
  • Tramblay, Y., , Ruelland D. , , Somot S. , , Bouaicha R. , , and Servat E. , 2013: High-resolution MED-CORDEX regional climate model simulations for hydrological impact studies: A first evaluation of the ALADIN-Climate model in Morocco. Hydrol. Earth Syst. Sci., 17, 37213739, doi:10.5194/hess-17-3721-2013.

    • Search Google Scholar
    • Export Citation
  • van Genuchten, M. T., 1980: A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Sci. Soc. Amer. J., 44, 892898, doi:10.2136/sssaj1980.03615995004400050002x.

    • Search Google Scholar
    • Export Citation
  • Vano, J. A., , Das T. , , and Lettenmaier D. P. , 2012: Hydrologic sensitivities of Colorado River runoff to changes in precipitation and temperature. J. Hydrometeor., 13, 932949, doi:10.1175/JHM-D-11-069.1.

    • Search Google Scholar
    • Export Citation
  • Vautard, R., and et al. , 2013: The simulation of European heat waves from an ensemble of regional climate models within the EURO-CORDEX project. Climate Dyn., 41, 25552575, doi:10.1007/s00382-013-1714-z.

    • Search Google Scholar
    • Export Citation
  • Wigley, T. M. L., , and Jones P. D. , 1985: Influences of precipitation changes and direct CO2 effects on streamflow. Nature, 314, 149152, doi:10.1038/314149a0.

    • Search Google Scholar
    • Export Citation
Save