• Ahrens, B., 2006: Distance in spatial interpolation of daily rain gauge data. Hydrol. Earth Syst. Sci., 10 , 197208.

  • American Meteorological Society, 2004: AMS statement on meteorological drought. Bull. Amer. Meteor. Soc., 85 , 771773.

  • Ashraf, M., , Loftis J. C. , , and Hubbard K. G. , 1997: Application of geostatistics to evaluate partial weather station networks. Agric. For. Meteor., 84 , 255271.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bacanli, U. G., , Firat M. , , and Dikbas F. , 2009: Adaptive Neuro-Fuzzy Inference System for drought forecasting. Stochastic Environ. Res. Risk Assess., 23 , 11431154.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bonaccorso, B., , Bordi I. , , Cancelliere A. , , Rossi G. , , and Sutera A. , 2003: Spatial variability of drought: An analysis of the SPI in Sicily. Water Resour. Manage., 17 , 273296.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chang, C. L., , Lo S. L. , , and Yu S. L. , 2005: Applying fuzzy theory and genetic algorithm to interpolate precipitation. J. Hydrol., 314 , 92104.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chartzoulakis, K., , and Psarras G. , 2005: Global change effects on crop photosynthesis and production in Mediterranean: The case of Crete, Greece. Agric. Ecosyst. Environ., 106 , 147157.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chartzoulakis, K., , Paranychianakis N. V. , , and Angelakis A. N. , 2001: Water resources management in the island of Crete, Greece, with emphasis on the agricultural use. Water Policy, 3 , 193205.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cheng, C.-H., 1998: A new approach for ranking fuzzy numbers by distance method. Fuzzy Sets Syst., 95 , 307317.

  • Christensen, J. H., , and Christensen O. B. , 2007: A summary of the PRUDENCE model projections of changes in European climate by the end of this century. Climatic Change, 81 , 730. doi:10.1007/s10584-006-9210-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Christensen, J. H., , Boberg F. , , Christensen O. B. , , and Lucas-Picher P. , 2008: On the need for bias correction of regional climate change projections of temperature and precipitation. Geophys. Res. Lett., 35 , L20709. doi:10.1029/2008GL035694.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Christensen, J. H., , Kjellström E. , , Giorgi F. , , Lenderink G. , , and Rummukainen M. , 2010: Weight assignment in regional climate models. Climate Res., 44 , 179194.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Christensen, O. B., , Drews M. , , Christensen J. H. , , Dethloff K. , , Ketelsen K. , , Hebestadt I. , , and Rinke A. , 2006: The HIRHAM Regional Climate Model Version 5 (β). Tech Rep 06-17. ISSN 1399-1388. DMI, Copenhagen.

    • Search Google Scholar
    • Export Citation
  • Collins, M., , Booth B. B. B. , , Bhaskaran B. , , Harris G. R. , , Murphy J. M. , , Sexton D. H. M. , , and Webb M. J. , 2009: Climate model errors, feedbacks and forcings: A comparison of perturbed physics and multi-model ensembles. Climate Dyn., 27 , 127147. doi:10.1007/ s00382-010-0808-0.

    • Search Google Scholar
    • Export Citation
  • Dirks, K. N., , Hay J. E. , , Stow C. D. , , and Harris D. , 1998: High-resolution studies of rainfall on Norfolk Island. Part II: Interpolation of rainfall data. J. Hydrol., 208 , 187193.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • EUR-Lex, cited. 2000: Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for community action in the field of water policy. [Available online at http://ec.europa.eu/environment/water/water-framework/index_en.html].

    • Search Google Scholar
    • Export Citation
  • Founda, D., , and Giannakopoulos C. , 2009: The exceptionally hot summer of 2007 in Athens, Greece—A typical summer in the future climate? Global Planet. Change, 67 , 227236. doi:10.1016/j.gloplacha.2009.03.013.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gaiser, T., and Coauthors, 2008: Development of a regional model for integrated management of water resources at the basin scale. Phys. Chem. Earth, 33 , 175182. doi:10.1016/j.pce.2007.04.018.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Giorgi, F., , and Mearns L. O. , 1999: Introduction to special section: Regional climate modelling revisited. J. Geophys. Res., 104 , 63356352.

  • Hayes, M. J., , Svoboda M. D. , , Wilhite D. A. , , and Vanyarkho O. V. , 1999: Monitoring the 1996 drought using the standardized precipitation index. Bull. Amer. Meteor. Soc., 80 , 429438.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ines, A. V. M., , and Hansen J. W. , 2006: Bias correction of daily GCM rainfall for crop simulation studies. Agric. For. Meteor., 138 , 4453. doi:10.1016/j.agrformet.2006.03.009.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jacob, D., 2001: A note to the simulation of the annual and inter-annual variability of the water budget over the Baltic Sea drainage basin. Meteor. Atmos. Phys., 77 , 6173.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jacob, D., and Coauthors, 2007: An inter-comparison of regional climate models for Europe: Design of the experiments and model performance. Climatic Change, 81 , (Suppl.). 3152.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jacob, D., , Christensen O. B. , , Doblas-Reyes F. J. , , Goodess C. , , Tank A. K. , , Lorenz P. , , and Roeckner E. , 2008: Information on observations, global and regional modelling data availability and statistical downscaling. ENSEMBLES Tech. Rep. 4, 10 pp. [Available online at http://ensembles-eu.metoffice.com/tech_reports/ETR_4_vn1.pdf].

    • Search Google Scholar
    • Export Citation
  • Jaeger, E. B., , Anders I. , , Lüthi D. , , Rockel B. , , Schär C. , , and Seneviratne S. , 2008: Analysis of ERA40-driven CLM simulations for Europe. Meteor. Z., 17 , 349367.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kjellström, E., , Bärring L. , , Gollvik S. , , Hansson U. , , Jones C. , , Samuelsson P. , , Rummukainen M. , , Ullerstig A. , , Willén U. , , and Wyser K. , 2005: A 140-year simulation of European climate with the new version of the Rossby Centre regional atmospheric climate model (RCA3). SMHI Reports Meteorology and Climatology, 108 .Swedish Meteorological and Hydrological Institute, Norrköping, Sweden, 54pp.

    • Search Google Scholar
    • Export Citation
  • Kidson, J. W., , and Thompson C. S. , 1998: A comparison of statistical and model-based downscaling techniques for estimating local climate variations. J. Climate, 11 , 735753.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Komuscu, A. U., 1999: Using the SPI to analyze spatial and temporal patterns of drought in Turkey. Drought Network News, 11 , 713.

  • Koutroulis, A. G., , and Tsanis I. K. , 2010: A method for estimating flash flood peak discharge in a poorly gauged basin: Case study for the 13–14 January 1994 flood, Giofiros basin, Crete, Greece. J. Hydrol., 385 , 150164. doi:10.1016/j.jhydrol.2010.02.012.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Larsén, X. G., , Mann J. , , Göttel H. , , and Jacob D. , 2008: Wind climate and extreme winds from the regional climate model REMO. Proc. European Wind Energy Conf. and Exhibition, Brussels, Belgium, EWEA, 58–62.

    • Search Google Scholar
    • Export Citation
  • Law, A. M., , and Kelton W. D. , 1982: Simulation Modeling and Analysis. McGraw-Hill, 400 pp.

  • Loukas, A., , and Vasiliades L. , 2004: Probabilistic analysis of drought spatiotemporal characteristics in Thessaly region, Greece. Nat. Hazards Earth Syst. Sci., 4 , 719731.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Manios, T., , and Tsanis I. K. , 2006: Evaluating water resources availability and wastewater reuse importance in the water resources management of small Mediterranean municipal districts. Resour. Conserv. Recycl., 47 , 245259.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • McKee, T. B., , Doesken N. J. , , and Kleist J. , 1993: The relationship of drought frequency and duration to time scales. Proc. Eighth Conf. on Applied Climatology, Anaheim, CA, Amer. Meteor. Soc., 179–184.

    • Search Google Scholar
    • Export Citation
  • Meyer, S. J., , Hubbard K. G. , , and Wilhite D. A. , 1993: A crop-specific drought index for corn: I. Model development and validation. Agron. J., 85 , 388395.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Murphy, J., 1999: An evaluation of statistical and dynamical techniques for downscaling local climate. J. Climate, 12 , 22562284.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mylopoulos, Y. A., , and Kolokytha E. G. , 2008: Integrated water management in shared water resources: The EU Water Framework Directive implementation in Greece. Phys. Chem. Earth, 33 , 347353. doi:10.1016/j.pce.2008.02.031.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Naoum, S., , and Tsanis I. K. , 2004: Orographic precipitation modeling with multiple linear regression. J. Hydrol. Eng., 9 , 79102.

  • Palmer, D. J., , Höck B. K. , , Kimberley M. O. , , Watt M. S. , , Lowe D. J. , , and Payn T. W. , 2009: Comparison of spatial prediction techniques for developing Pinus radiata productivity surfaces across New Zealand. For. Ecol. Manage., 258 , 20462055. doi:10.1016/j.foreco.2009.07.057.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Palmer, W., 1965: Meteorological drought. U.S. Weather Bureau Research Paper 45, 58 pp.

  • Rossi, G., 2000: Drought mitigation measures: A comprehensive framework. Drought and Drought Mitigation in Europe, J. Voght and F. Somma, Eds., Kluwer Academic Publishers, 233–246.

    • Search Google Scholar
    • Export Citation
  • Rossi, G., , Benedini M. , , Tsakiris G. , , and Giakoumakis S. , 1992: On regional drought estimation and analysis. Water Resour. Manage., 6 , 249277.

  • Shepard, D., 1968: A two-dimensional interpolation function for irregularly spaced data. Proc. 23rd ACM National Conf., Princeton, NJ, Association of Computing Machinery, 517–524, doi:10.1145/800186.810616.

    • Search Google Scholar
    • Export Citation
  • Siska, P. P., , and Hung I.-K. , 2001: Assessment of kriging accuracy in the GIS environment. Proc. 21st Annual ESRI Int. User Conf., San Diego, CA, Environmental Systems Research Institute, Paper 280.

    • Search Google Scholar
    • Export Citation
  • Solomon, S., , Qin D. , , Manning M. , , Marquis M. , , Averyt K. , , Tignor M. M. B. , , Miller H. L. Jr., , and Chen Z. , Eds. 2007: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.

    • Search Google Scholar
    • Export Citation
  • Sönmez, F. K., , Kömüscü , , Erkan A. , , and Turgu E. , 2005: An analysis of spatial and temporal dimension of drought vulnerability in Turkey using the standardized precipitation index. Nat. Hazards, 35 , 243264. doi:10.1007/s11069-004-5704-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tabios, G. Q., , and Salas J. D. , 1985: A comparative analysis of techniques for spatial interpolation of precipitation. J. Amer. Water Resour. Assoc., 21 , 365380.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tannehill, I. R., 1947: Drought—Its Causes and Effects. Princeton University Press, 597 pp.

  • Teegavarapu, R. S. V., , Tufail M. , , and Ormsbee L. , 2009: Optimal functional forms for estimation of missing precipitation data. J. Hydrol., 374 , 106115.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tsakiris, G., , and Vangelis H. , 2004: Towards a drought watch system based on spatial SPI. Water Resour. Manage., 18 , 112.

  • Tsakiris, G., , Pangalou D. , , Tigkas D. , , and Vangelis H. , 2007: Assessing the areal extent of drought. Proc. EWRA Int. Symp. on Water Resource Management: New Approaches and Technologies, Chania, Greece, European Water Resources Association, 59–66.

    • Search Google Scholar
    • Export Citation
  • Tsanis, I. K., , and Naoum S. , 2003: The effect of spatially distributed meteorological parameters on irrigation water demand assessment. Adv. Water Res., 26 , 311324.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Van der Linden, P., , and Mitchell J. F. B. , Eds. 2009: ENSEMBLES: Climate change and its impacts: Summary of research and results from the ENSEMBLES project. Met Office Hadley Centre, 160 pp. [Available from Met Office Hadley Centre, FitzRoy Road, Exeter EX1 3PB, United Kingdom].

    • Search Google Scholar
    • Export Citation
  • Van Meijgaard, E., , van Ulft L. H. , , van de Berg W. J. , , Bosveld F. C. , , van den Hurk B. J. J. M. , , Lenderink G. , , and Siebesma A. P. , 2008: The KNMI regional atmospheric climate model RACMO, version 2.1. KNMI Publ. TR-302, KNMI, De Bilt, pp. 50.

    • Search Google Scholar
    • Export Citation
  • Ware, C., , Knight W. , , and Wells D. , 1991: Memory intensive statistical algorithms for multibeam bathymetric data. Comput. Geosci., 17 , 985993.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wei, T. C., , and McGuinness J. L. , 1973: Reciprocal Distance Squared Method: A computer technique for estimating area precipitation. U.S. Agricultural Research Service Tech. Rep. ARS-Nc-8., 29 pp.

    • Search Google Scholar
    • Export Citation
  • Wilby, R. L., , Hay L. E. , , Gutowski W. J. Jr., , Arritt R. W. , , Takle E. S. , , Pan Z. , , Leavesley G. H. , , and Clark M. P. , 2000: Hydrological responses to dynamically and statistically downscaled climate model output. Geophys. Res. Lett., 27 , 11991202.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wilhite, D. A., 2000: Drought as a natural hazard: Concepts and definitions. Drought: A Global Assessment, Natural Hazards and Disasters Series, Vol. 2, Routledge Publishers, 3–18.

    • Search Google Scholar
    • Export Citation
  • Wood, A. W., , Maurer E. P. , , Kumar A. , , and Lettenmaier D. P. , 2002: Long-range experimental hydrologic forecasting for the eastern United States. J. Geophys. Res., 107 , 4429. doi:10.1029/2001JD000659.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wu, H., , Svoboda M. D. , , Hayes M. J. , , Wilhite D. A. , , and Wen F. , 2007: Appropriate application of the Standardized Precipitation Index in arid locations and dry seasons. Int. J. Climatol., 27 , 6579.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yamoah, C. F., , Walters D. T. , , Shapiro C. A. , , Francis C. A. , , and Hayes M. J. , 2000: Standardized precipitation index and nitrogen rate effects on crop yields and risk distribution in maize. Agric. Ecosyst. Environ., 80 , 113120.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • View in gallery

    Hypothetical region consisting of A, B, and C watersheds.

  • View in gallery

    Monthly precipitation time series for the three hypothetical watersheds of Fig. 1.

  • View in gallery

    Monthly SPI 48 (48-month time scale) time series for the 3 hypothetical watersheds.

  • View in gallery

    Area of study, location of watersheds of Tavronitis, Geropotamos, and Petras, and ENSEMBLES grid over Crete Island.

  • View in gallery

    Elevation map of Crete Island.

  • View in gallery

    Mean annual precipitation time series for Tavronitis, Geropotamos, and Petras watersheds for the period 1973–2004.

  • View in gallery

    Average annual precipitation at watershed level for the 1973–2004 period and location of precipitation stations over Crete.

  • View in gallery

    Monthly SN–SPI 48 (48-month time scale) time series for the 3 hypothetical watersheds.

  • View in gallery

    The SN–SPI sensitivity to the precipitation variability of dry basin A for basin (a) A, (b) B, and (c) C.

  • View in gallery

    As in Fig. 9, but for wet basin C.

  • View in gallery

    The (top) 12-, (middle) 24-, and (bottom) 48-month time scale SPI for (a) Tavronitis, (b) Geropotamos, and (c) Petras for 1973–2004.

  • View in gallery

    As in fig. 11, but for the SPI and SN–SPI time series.

  • View in gallery

    As in Fig. 12, but for SPI and SN–SPI time series and only Tavronitis and Petras.

  • View in gallery

    Drought extremity map in terms of time percentage for 1973–2004 for 12-, 24-, and 48-month SN–SPI.

  • View in gallery

    As in Fig. 14, but for the SPI.

  • View in gallery

    Crete percentage area that experiences drought conditions (−1 < SN–SPI ≤ 0; −1,5 < SN–SPI ≤ −1; −2 < SN–SPI ≤ −1,5; and SN–SPI ≤ −2—mildly dry, moderately dry, severely dry, and extremely dry, respectively) for 1973–2004 for (top to bottom) the 12-, 24-, and 48-month SN–SPI.

  • View in gallery

    The 12-, 24-, and 48-month (a) SN–SPI and (b) SPI for Tavronitis, Geropotamos, and Petras for 2010–2100.

  • View in gallery

    As in Fig. 16, but for 2010–2100.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 153 153 15
PDF Downloads 135 135 16

Spatiotemporal Characteristics of Meteorological Drought for the Island of Crete

View More View Less
  • 1 Technical University of Crete, Chania, Greece
© Get Permissions
Full access

Abstract

A modified drought index, named the spatially normalized–standardized precipitation index (SN-SPI), has been developed for assessing meteorological droughts. The SN–SPI is a variant index to the standardized precipitation index and is based on the probability of precipitation at different time scales, but it is spatially normalized for improved assessment of drought severity. Results of this index incorporate the spatial distribution of precipitation and produce improved drought warnings. This index is applied in the island of Crete, Greece, and the drought results are compared to the ones of SPI. A 30-year-long average monthly precipitation dataset from 130 watersheds of the island is used by the above indices for drought classification in terms of its duration and intensity. Bias-adjusted monthly precipitation estimates from an ensemble of 10 regional climate models were used to quantify the influence of global warming to drought conditions over the period 2010–2100. Results based on both indices (calculated for three time scales of 12, 24, and 48 months) from 3 basins in west, central, and east parts of the island show that 1) the extreme drought periods are the same (reaching 7% of time) but the intensities based on SN–SPI are lower; 2) the area covered by extreme droughts is 3% (first time scale), 16% (second time scale), and 25% (third time scale), and 96% (first time scale), 95% (second time scale), and 80% (third time scale) based on the SN–SPI and SPI, respectively; 3) concerning the longest time scale (48 months), more than half of the area of Crete is about to experience drought conditions during 28%, 69%, and 97% for 2010–40, 2040–70, and 2070–2100, respectively; and 4) extremely dry conditions will cover 52%, 33%, and 25% of the island for the future 90-year period using 12-, 24-, and 48-month SN–SPI, respectively.

Corresponding author address: Ioannis Tsanis, Technical University of Crete, Department of Environmental Engineering, University Campus, Kounoupidiana, Chania, Crete P.C. GR 73100, Greece. Email: tsanis@hydromech.gr

This article included in the Water and Global Change (WATCH) special collection.

Abstract

A modified drought index, named the spatially normalized–standardized precipitation index (SN-SPI), has been developed for assessing meteorological droughts. The SN–SPI is a variant index to the standardized precipitation index and is based on the probability of precipitation at different time scales, but it is spatially normalized for improved assessment of drought severity. Results of this index incorporate the spatial distribution of precipitation and produce improved drought warnings. This index is applied in the island of Crete, Greece, and the drought results are compared to the ones of SPI. A 30-year-long average monthly precipitation dataset from 130 watersheds of the island is used by the above indices for drought classification in terms of its duration and intensity. Bias-adjusted monthly precipitation estimates from an ensemble of 10 regional climate models were used to quantify the influence of global warming to drought conditions over the period 2010–2100. Results based on both indices (calculated for three time scales of 12, 24, and 48 months) from 3 basins in west, central, and east parts of the island show that 1) the extreme drought periods are the same (reaching 7% of time) but the intensities based on SN–SPI are lower; 2) the area covered by extreme droughts is 3% (first time scale), 16% (second time scale), and 25% (third time scale), and 96% (first time scale), 95% (second time scale), and 80% (third time scale) based on the SN–SPI and SPI, respectively; 3) concerning the longest time scale (48 months), more than half of the area of Crete is about to experience drought conditions during 28%, 69%, and 97% for 2010–40, 2040–70, and 2070–2100, respectively; and 4) extremely dry conditions will cover 52%, 33%, and 25% of the island for the future 90-year period using 12-, 24-, and 48-month SN–SPI, respectively.

Corresponding author address: Ioannis Tsanis, Technical University of Crete, Department of Environmental Engineering, University Campus, Kounoupidiana, Chania, Crete P.C. GR 73100, Greece. Email: tsanis@hydromech.gr

This article included in the Water and Global Change (WATCH) special collection.

1. Introduction

One of the least understood phenomena connected to the weather that affects environment, society, and economy in large areas all over the world is drought (Rossi et al. 1992). In the literature, many definitions of drought are given but the concept of a water deficit is the keynote in every definition: “Drought is a random condition of severe reduction of water supply availability (compared to normal value) extending along a significant period of time over a large region” (Rossi 2000). All points of view seem to agree that drought is characterized by a significant decrease of water availability caused by a deficit in precipitation during a significant period over a large area. The effects and impacts of drought often accumulate slowly over a considerable period of time and may linger for years after the termination of the event, so there are always lags in perceiving these effects and impacts. Drought is an objective phenomenon, and so is its onset and end. The main issue is how to describe and how to quantify them. Because of this, drought is often referred to as a “creeping phenomenon” (Tannehill 1947).

There can be significant variations of drought impacts between regions due to the different characteristics of the economy, society, and the environment. Hence, different types of droughts are identified (American Meteorological Society 2004). Meteorological drought is defined usually by the departure of precipitation from the “normal” or average amount and the duration of the dry period, while agricultural drought refers to situations in which the soil moisture is no longer sufficient to meet the needs of the crops growing in the area, focusing on properties such as precipitation shortage, differences between potential and actual evapotranspiration, and soil moisture deficits. Finally, hydrological drought associates the effect of periods of precipitation shortfalls on surface or subsurface water supply (Wilhite 2000). In spite of the fact that precipitation is the primary factor that controls drought, other factors such as high temperature or dry winds contribute to the amplification of its intensity. The severity of drought depends on the moisture deficit degree, the duration of the phenomenon, and its spatial extent. Drought impacts first appear on agriculture, which is prone to be affected by soil moisture decrease and high evapotranspiration. During extended dry periods, soil water depletes fairly rapidly. On the other hand, the last water resources to be affected by an extended dry period are usually surface and subsurface (Sönmez et al. 2005).

Drought indices are indispensible tools to detect, monitor, and evaluate drought events in both time and space. A large number of studies are included in the international literature on testing the efficiency and the effectiveness of various drought indices regarding detection and monitoring drought events and regional drought analysis (Palmer 1965; McKee et al. 1993; Meyer et al. 1993). The standardized precipitation index (SPI; McKee et al. 1993) and the Palmer drought severity index (PDSI; Palmer 1965) are the most commonly used amongst the drought indices. As a matter of fact, drought indices contain a large amount of data on rainfall, streamflow, snow, and other indicators that transform these huge datasets into a comprehensible picture. Hence, a drought index value is typically a single number—more useful than a raw dataset for decision making.

The application of the SPI covers a significant part of many studies that have been carried out over the last decades (Bonaccorso et al. 2003; Loukas and Vasiliades 2004; Wu et al. 2007). Bacanli et al. (2009) assumed that when the analysis period (time scale) increases, drought is observed less but lasts longer. According to their study, the SPI showed short-term conditions with seasonal variation for 3- and 6-month periods, while 9- and 12-month periods used showed drought with average duration; long-term drought is assessed when using a 24-month period. It is also possible to provide a comprehensive picture for the magnitude, duration, and spatial character of drought, as established by Tsakiris and Vangelis (2004).

The analysis of short temporal scales (less than 12 months), indicative of agricultural drought (Yamoah et al. 2000), is of little interest because of decreased economic importance of nonirrigated agriculture of the study region. Longer temporal units (over 12 months) are used in the present study since they are better suited to monitor hydrological than agricultural droughts (Hayes et al. 1999; Komuscu 1999).

Although SPI is widely used for assessing drought occurrence, there are some limitations in providing relative information when applied for different regions (at river basin scale). For example, three watersheds A, B, and C (Fig. 1) are considered with 720-, 1320-, and 1920-mm average annual rainfall, respectively, for the long-term period of 1970–2000. This different precipitation distribution over the area could be due to the higher elevation of basin C (mountainous), medium elevation of basin B, and lower elevation of basin A (flat), or due to regional atmospheric patterns that dominantly derive from the northwest and head toward the southeast, passing over the area and delivering a higher amount of orogenic precipitation over the north and northwest part of the area (basins C and B) and lower precipitation over basin A, influenced by the “rain shadow” effect. For this hypothetical concept we assume a common fluctuation of precipitation over the three watersheds, presented in Fig. 2. Watersheds A, B, and C can be characterized as dry, normal, and wet watersheds, respectively. Monthly precipitation time series of basins B and C results from a simple addition of 50 and 100 mm per month to the time series of basin A.

The SPI 48 monthly time series of (48-month time scale) are presented in Fig. 3. These series, presenting the departure of precipitation from the median value for each watershed, are practically the same. Here lies the limitation of the SPI when used to provide relative information among basins or region-scale areas. The severity of the 1990–95 period is identical for the 3 basins, although basin C receives much more precipitation than basins B and A. For example, the SPI 48 value for September 1993 is −2.4 for all the watersheds, despite the fact that watersheds C and B received 3 and 2 times more precipitation, respectively, during the corresponding 48-month previous period (October 1989–September 1993). This example presents the shortcoming of the SPI in providing relative information related to the water resources of a region. Although the period 1990–95 is severely dry for each of the 3 hypothetical basins, water excess from wet watershed C could be used to cover the water shortage of dry watershed A.

The recently adopted European Water Framework Directive (WFD) 2000/60/EC (EUR-Lex 2000) establishes a new institutional framework, giving directions for the common approach and objectives, principals, definitions, and measures for the management of waters in Europe (Mylopoulos and Kolokytha 2008). The central feature of the WFD is the use of river basins as the basic unit for all planning and management actions (Gaiser et al. 2008). This basin-scale approach was adopted for the present study, attempting to improve the information and assessment of drought for the purpose of providing more robust integrated water management insights under the implementation of the WFD.

The objective of this study is to introduce a modified SPI drought index—a variant of the SPI: the spatially normalized–standardized precipitation index (SN–SPI) allows the comparison between watersheds with different mean annual precipitation. Moreover, the paper intends to ascertain the spatiotemporal character of drought occurrence in Crete and examine potential future drought patterns based on results from regional climate models.

2. Study area and data description

The island of Crete is located in the southeastern part of the Mediterranean region (Fig. 4) and is well known to be one of the most drought-prone areas of Greece. Political interests and disputes among the four prefectures and the more than 100 municipalities of the island, as well as poor water management, have created a public belief that water resources are inadequate and that some kind of drought is imminent (Manios and Tsanis 2006).

The island has a surface area of 8336 km2, the mean altitude is 460 m, and its population reaches 600 000 people. The landscape of Crete agrees with the general pattern of Greek landscape that consists of mountainous terrain, since the island is the continuation of the mountain ranges of the Greek mainland.

Crete consists of four prefectures—from west to east: Chania, Rethimnon, Heraklio, and Lassithi. The mean annual precipitation is estimated to be 750 mm, varies from east [440 mm (Ierapetra, elevation 10 m)] to west [2118 mm (Askifou, elevation 740 m)], and the potential renewable water resources reach 2650 mm3. In spite of the fact that the annual precipitation is rather high, it is estimated that about 63% evapotranspirates, 10% outflows to the sea, and only 27% recharges the groundwater (Table 1). The actual water use is about 485 mm3 year−1. The main water use in Crete covers irrigation, with a high percentage of 83.3% of the total consumption. Domestic use, including tourism, covers 15.6% and industrial use 1% of the total consumption (Chartzoulakis et al. 2001). In Crete, there are significant regional variations in water availability. The eastern and southern parts are more arid than the western and northern parts, as there is higher precipitation in the northwestern coastal areas and lower in the southeastern part of the island (Chartzoulakis and Psarras 2005). Western Crete receives higher amounts and rates of precipitation than eastern Crete (Naoum and Tsanis 2004). This is partly because of the regional atmospheric patterns that dominantly derive from the northwest and head toward the southeast. Another important factor is the morphological variability that presents higher elevation and steepest slopes in the west part of the island (Fig. 5), where orographic effects tend to increase both frequency and intensity of precipitation (Naoum and Tsanis 2004; Koutroulis and Tsanis 2010). The uneven spatial and temporal precipitation distributions of Crete, although common in many Mediterranean areas, have significant impact when they are compounded by the water demands associated with intensive agricultural activities and the tourism industry (Tsanis and Naoum 2003).

The climate ranges between subhumid Mediterranean and semiarid with long hot and dry summer and relatively humid and cold winter. In the coastal zone, extremely low temperatures are very rare during winter. Also, temperature decreases by altitude during winter, while summer is characterized by temperature increase by altitude from the coast to the mainland plains.

Several studies have been performed for the application of the SPI. Especially in the area of Crete, Tsakiris and Vangelis (2004) concluded that the eastern part of the island suffers more frequently from droughts according to a method based on the estimation of the SPI and its use for characterizing drought. A digital terrain model based on spatial distribution utilizing a grid analysis and a simple computer calculating process was used, and it was deduced that the proposed procedure could be easily applied to an area of mesoscale dimensions. It was also concluded that a significantly persistent drought occurrence was noted during the period 1987–94, while distinct drought events were observed in the years 1973–74, 1976–77, 1985–86, and 1999–2000. Additionally, Tsakiris et al. (2007) estimated drought areal extent for eastern Crete using the SPI and reconnaissance drought index (RDI), and deduced that the driest year during the examined period from 1962–63 to 1991–92 is 1989–90.

Monthly precipitation data were compiled by the Water Resources Department of the Prefecture of Crete (WRDPC) service for 67 precipitation stations. The stations mainly cover the eastern part of the island, which has a higher level of agricultural and tourism activity than the western part. Out of the entire dataset, 14 gauges recorded only for 4–10 years and were excluded from the analysis, while the rest had sufficient measurements for over 30 years and were used for the identification of drought. The gauges were located at elevations that ranged from sea level (the prefecture of Iraklion, central Crete) to 905 m MSL (the prefecture of Lasithi, eastern Crete) (Chartzoulakis et al. 2001).

In simple terms, the monthly precipitation data were measured, interpolated, and monthly precipitation time series data of 130 watersheds were generated. The proposed methodology was applied to the 130 major basins of the island of Crete, and for illustration purposes, three test basins from the western, middle, and eastern parts of the island are presented (Fig. 4). These data cover a 30-year time period for each month of the hydrological year (September–August) from 1973 to 2004. The mean annual precipitation time data for the three representative watersheds are summarized in Table 2 and illustrated in the diagram of Fig. 6.

The results of a climate change scenario using observed greenhouse gas (GHG) concentrations until 2000 and Special Report on Emissions Scenarios (SRES) A1B concentrations (Solomon et al. 2007) until 2100 [carried out within the ENSEMBLE-based Predictions of Climate Changes and their Impacts (ENSEMBLES) project (Van der Linden and Mitchell 2009)] were produced by a set of 10 regional climate models for the period 1970–2100. Precipitation time series of the 25-km mesh resolution (Fig. 4) of ENSEMBLES dataset were aggregated to the 130 major basins of the island of Crete using areal weighting factors.

Observed daily precipitation time series data [obtained from 53 rainfall stations via the inverse distance weighted (IDW) method] for the period 1973–2000 were used to correct the bias of the regional climate model (RCM) results for the 130 watersheds—that is, compensate for differences in monthly mean and variance without precluding future changes in variability. RCM rainfall was adjusted in order to approximate the long-term frequency and intensity distribution observed at each basin. The SPI and SN–SPI for each watershed was then calculated for the past (baseline: 1973–2000) and future periods.

3. Methodology

a. The SPI

The SPI seems to win universal applicability. In the present paper, the SPI is used for assessing drought occurrence in Crete. The index offers the capability to assess drought conditions over a wide range of time scales, while comparison between dry and wet periods on different locations is permitted. Moreover, it is based on precipitation alone, so that a drought could be assessed even when other hydrometeorological data are not available (Bonaccorso et al. 2003).

There is a general agreement about the fact that the SPI computed on shorter time scales (3 or 6 months) describes drought events that affect agricultural practices, while on the longer ones (12, 24, or 48 months), the effects of a precipitation deficit on different water resource components are given (soil moisture, streamflow, groundwater, and reservoir storage). This study presents the results of long-period drought (48 months) analysis.

The SPI was developed by McKee et al. (1993). In its original version, precipitation for a long period at a station is fitted to a Gamma probability distribution, which is then required to be transformed into a normal distribution so that the mean SPI value is zero. The index values are then the standardized deviations of the transformed precipitation totals from the mean. The gamma distribution is defined by its frequency or probability density function:
i1525-7541-12-2-206-e1
where α is a shape parameter (α > 0), β is a scale parameter (β > 0), x is the precipitation amount (x > 0), and Γ(α) is the gamma function.

Positive SPI values denote greater than median precipitation whereas negative values denote less than median precipitation. Periods with drought conditions are represented by relatively high negative deviations. Specifically, the “drought” part of the SPI range is arbitrary divided in four categories: mildly dry (0 > SPI > −0.99), moderately dry (−1.0 > SPI > −1.49), severely dry (−1.5 > SPI > −1.99), and extremely dry conditions (SPI < −2.0). A drought event is considered to start when SPI reaches negative values and ends when SPI becomes positive again (McKee et al. 1993). Thresholds of the SPI for drought characterization are given in Table 3.

b. The IDW method

The IDW method (Wei and McGuinness 1973) is most commonly used for estimating missing environmental data. IDW has been widely used in many different fields, such as hydrology, earth science (Ware et al. 1991; Ashraf et al. 1997; Cheng 1998), etc., and is commonly applied to estimate average precipitation and interpolate unknown rainfall (Chang et al. 2005). Even though IDW is a dated method, recent techniques for precipitation interpolation (e.g., Teegavarapu et al. 2009) have not managed to dramatically improve its results. Palmer et al. (2009) compared ordinary kriging (OK), regression kriging (RK), and IDW interpolation techniques, resulting in similar performance rankings and providing predictions of similar precision and bias in their study (sparse and distant observations). Assessments of uncertainty associated with interpolation techniques available in most GIS packages suggest that kriging, IDW, Thiessen polygons, and triangulated irregular network (TIN) interpolations performed almost on the same level (Siska and Hung 2001). Furthermore, regression modeling and kriging techniques require good judgment, experience, and expertise by the practitioner, compared with IDW and its more rudimentary approach (Palmer et al. 2009). Undoubtedly, it can be shown that statistical interpolation methods like multiple linear regression, optimal interpolation, or kriging can perform better than IDW, but only if the data’s density is sufficient (Ahrens 2006). While existing geostatistical methods like kriging have been known to provide better results for the spatial interpolation of precipitation (Tabios and Salas 1985), the degree of complexity and the computational effort they require do not justify their use for high-resolution networks (Dirks et al. 1998).

The spatial conversion of point precipitation data was driven by the IDW method, which includes multivariate interpolation—a process of assigning values to unknown points by using values usually from a scattered set of known points (Shepard 1968). A general form of finding an interpolated value u for a given point x using IDW is an interpolating function:
i1525-7541-12-2-206-e2
i1525-7541-12-2-206-e3
where x denotes an interpolated (arbitrary) point, xk is an interpolating (known) point, d is a given distance (metric operator) from the known point xk to the unknown point x, N is the total number of known points used in interpolation, and p is a positive real number, called the power parameter.

For the present study, the IDW method was used to interpolate daily precipitation data from 53 stations over Crete island (Fig. 7) in an ArcView GIS package because of the partially sparse resolution of the gauging network (especially over the Western part of the island) and the low computational effort of the method. Daily interpolated precipitation grids were averaged at basin scale and aggregated at a monthly time step for further SPI and SN–SPI calculations.

c. The SN–SPI

The objective of the new modified SPI is the potent comparison of drought events among different areas with different mean annual precipitation at different times. The procedure includes the normalization of SPI values through the incorporation of the precipitation values. The calculation of the SN–SPI is based on a two-step procedure. The first step is the normalization of the SPI according to the relative average precipitation, based on a set of coefficients (ai, bi) that satisfy
i1525-7541-12-2-206-e4
i1525-7541-12-2-206-e5
where Pi is the mean monthly precipitation for each watershed i, and Pall the mean monthly precipitation for all watersheds. Given ai and bi, SPI′i for each watershed i is calculated through
i1525-7541-12-2-206-e6
i1525-7541-12-2-206-e7
With the above procedure SPIi, time series of each watershed i are modified properly in order to include the information of relative average precipitation among all the watersheds of the study area, resulting to the corresponding SPI′i time series. The second step includes the rescaling of SPI′i in order to meet the scale of SPIi, based on the coefficients c and d estimated through
i1525-7541-12-2-206-e8
i1525-7541-12-2-206-e9
where max(SPIi) is the maximum SPI value of all watersheds, min(SPIi) is the minimum SPI value of all watersheds, max(SPI′i) the maximum modified SPI value of all watersheds, and min(SPI′i) the minimum modified SPI value of all watersheds. Given c and d, the SN–SPI calculation is defined by
i1525-7541-12-2-206-e10
i1525-7541-12-2-206-e11
thus results of SN–SPI have the same scale of original SPI and can be easily compared.

Based on the example of the introduction with the three hypothetical basins, the corresponding SN–SPI time series were calculated. The application of the SN–SPI modifies the results of SPI, including the relative information of precipitation among the three hypothetical basins, as shown in Fig. 8 when compared with Fig. 3. The comparison of the dry, wet, or normal conditions of the basins is more representative through SN–SPI in terms of water resources, depicting the strength of the proposed index. The SN–SPI is based on the spatial variability of the precipitation. It is rather obvious that the sensitivity of precipitation variability is depicted on the index results, leading to some limitations on the application.

The sensitivity of the SN–SPI to the variability of basin-scale precipitation was examined based on the example of the three hypothetical watersheds A, B, and C (Fig. 1), with 720-, 1320-, and 1920-mm average annual rainfall, respectively, for the long-term period of 1970–2000. The temporal scale of 48-month SN–SPI was selected for sensitivity analysis. Similar results are obtained for smaller temporal scales.

  1. Sensitivity to the precipitation variability of the “dry” basin (A).By assuming different annual precipitation values for basin A, varying from −360 mm (−50%) to +360 mm (+50%) from the original 720 mm, SN–SPI datasets were estimated for basins A, B, and C as shown in Fig. 9. Higher precipitation for basin A (1080, 960, and 840 mm) results in higher SN–SPI values for the same basin and lower precipitation results in lower SN–SPI values for the wet period around 1981, as shown in Fig. 9a. The dry period (1990–96) values for basin A are not affected since the normalization of lower values is based on dry basin A. Since hypothetical basin A has average annual precipitation values ranging from 360 to 1080 mm that are always lower than basins B (1320 mm) and C (1920 mm), the value of precipitation in basin A determines the lower boundary of the SN–SPI values for all the cases of the sensitivity analysis and, thus, the whole procedure of estimating SN–SPI. Higher precipitation for basin A results in lower SN–SPI for basins B and C, and conversely, during the dry period 1990–96 (Figs. 9b,c). Generally, the basin with the lower precipitation modifies the intensity of SN–SPI of the dry spells for the “wetter” basins. The significant increase in precipitation in basin A (360–1080 mm) is depicted to the relative drought conditions of the wetter basins; this provides the appropriate information of the relative precipitation among these study basins through the modified index.
  2. Sensitivity to the precipitation variability of the “wet” basin (C).By assuming different annual precipitation values for basin C, varying from −360 mm (−18.75%) to +360 mm (+18.75%) from the original 1920 mm, SN–SPI datasets were estimated for basins A, B, and C as shown in Fig. 10. Higher precipitation for basin C (2280, 2160, and 2040 mm) results in lower SN–SPI values for dry basins A and B and lower precipitation results in higher SN–SPI values for the wet period around 1981, as shown in Figs. 10a,b. Higher precipitation for basin C results in lower SN–SPI for the same basin, and conversely, during the dry period 1990–96 (Fig. 10c). Generally, the basin with the higher precipitation modifies the intensity of SN–SPI of the wet spells for the “drier” basins.

d. Future climatic scenarios—The ENSEMBLE regional climate models dataset

RCM simulations provide useful insight into the climate conditions of a region, provided they can adequately reproduce the observed climate (Founda and Giannakopoulos 2009). A single-RCM approach of simulated climate is insufficient to provide the appropriate information for a comprehensive assessment of potential climate change and its impacts (Christensen and Christensen 2007; Jacob et al. 2007). Multimodel information has value, which can be enhanced with a performance-based weighting of the contributing models. Therefore, the drought scenarios were formulated by adopting the results of an ensemble of 10 RCMs within the framework of the European Union (EU) ENSEMBLES project (http://ensembles-eu.metoffice.com), regarding the climate change (precipitation data) in each area.

A regional climate model resolves small-scale atmospheric circulations and simulates atmospheric processes on diverse time scales using boundary conditions that are provided by global climate model simulations, observations, global analysis, and reanalysis products (Larsén et al. 2008). The ENSEMBLES RCMs were used to perform the downscaling of the results of global climate simulations to the regional scales. Thus, models were incorporated into the respective large-scale forcing, which could be given by global climate models (Van der Linden and Mitchell 2009; Jacob et al. 2008) using observed GHG concentrations until 2000 and SRES A1B concentrations until 2100.The ENSEMBLES domain includes continental Europe at 0.22° horizontal resolution.

Results of the SRES A1B climate change scenario (Solomon et al. 2007) over Europe for the period 1973–2100 was applied to basins in the island of Crete. The 10 RCMs (Table 4) used for precipitation analysis over Crete where chosen based on several criteria such as the correspondence of RCMs mesh, the forecast period, and the performance on their ability to simulate the present climate, calibrated for the 1973–2000 observed climate over Crete. RCM-specific weights were extracted in order to construct the optimal ensemble output for precipitation at a monthly time step and watershed level (Christensen et al. 2010). The ENSEMBLES RCM weights were based on a set of metrics that were defined by their ability to simulate the present climate, calibrated for the 1973–2000 observed precipitation, both for different months and for the Crete subregion at 25-km resolution (as adopted from D3.2.2 report of ENSEMBLES project). The metrics extracted for monthly precipitation are

  • F1: probability density distribution match of monthly precipitation analysis, and
  • F2: representation of the annual cycle in precipitation.

The combination of the performance of a particular RCM over the subregions and seasons is combined into a single weight for each RCM. This is done by a multiplication of the weights F1 and F2:
i1525-7541-12-2-206-e12
where all the individual weights have a value between 0 and 1. Here ni can be chosen as any positive number to weigh the various metrics differently (the value of 0 would imply equal weighting of the RCMs). The philosophy underlying this approach is that in order to receive a high weight, a model needs to perform well in all metrics considered so as to avoid to the extent possible the counterbalancing effect of different systematic biases. The final weighted aggregated values based on F1 and F2 metrics are presented in Table 4.

e. Bias correction of precipitation

Regional climate models tend to simulate precipitation to different statistical characteristics in relation to measured precipitation heights of different stations in the same cell. Bias correction constitutes the correction of the result exported from a model in such a manner that the results are statistically pertinent to the real data. The correction method varies according to the type and the temporal and spatial modes of the data. Some methods involve statistical characteristics recovery from the measured data and application over the model results, whereas some others are confined by a simple application of a correction factor over the model results. Numerous studies have demonstrated the need for bias correction in order to obtain useful results for applications in hydrology and water resources management (Murphy 1999; Kidson and Thompson 1998; Wilby et al. 2000; Christensen et al. 2008).

A two-step bias correction procedure was used that adjusts RCM rainfall to approximate the long-term frequency and intensity distribution observed at a given basin (Ines and Hansen 2006; Law and Kelton 1982; Wood et al. 2002). The correction involves truncating the RCM rainfall distribution at a basin level that approximately reproduces the long-term observed relative frequency of rainfall, then mapping the truncated RCM rainfall intensity distribution onto a gamma distribution fitted to observed intensity distribution.

4. Results and discussion

Short- and long-period characteristics represented by 12-, 24-, and 48-month time-scale values of SPI were calculated for three representative watersheds—Tavronitis (west), Geropotamos (central), and Petras (east)—aiming to provide an overview of prolonged drought occurrences during the period 1973–2004. The results of drought analysis in the territory of the island of Crete during this period show a clear tendency toward prolongation and greater severity of drought episodes (Fig. 11). Short-term SPI values for all watersheds provide a physically meaningful interpretation of seasonality. Fluctuations of the values are attributable to the seasonal character of precipitation, thereby highlighting the necessity of these time scales in events that affect agricultural activities. In 12-month SPI plots, the extreme drought of 1990 is present at all watersheds and an extra severe drought spell appears at Petras. As for 24-month SPI plots, severe and extreme drought characterize the period 1990–91 with various brief severe droughts at Geropotamos and Petras watersheds. Accordingly, analyzing in a long-term path the situation of the watersheds, the period 1988–96 was recorded to be a period of drought for Tavronitis, Geropotamos, and Petras. Extreme drought conditions were recorded in the years 1992–93 (Geropotamos) with the SPI reaching the value of −2.9. Drought phenomena were also observed during the period 2000–02. Severe drought spells were also recorded in the period of 1992–93 in Petras watershed. It is important to stress that in the Petras basin, the severe drought of 1992–93 is followed by a 9-year mild drought. In general, the southern and eastern parts of Crete suffer more from drought events.

SN–SPI is based on the spatial variability of the precipitation. It is rather obvious that the sensitivity of precipitation variability is depicted in the index results, leading to some limitations on the application. For example, a small watershed having extremely high precipitation would affect the estimates of the drought intensities of an entire region; and the drought intensities of the regions that receive low precipitation will be highly biased. Thus, a detailed analysis of the maximum and minimum precipitation values at watershed level, combined with the corresponding watershed area, should be performed before the application of the method. In the present study, the average area of the 130 catchments of Crete is 63.4 km2 with a maximum of 600 km2 and a minimum of 6 km2. Minimum annual precipitation of 518 mm corresponds to a basin of 85 km2, maximum annual precipitation corresponds to a basin of 160 km2, and average annual precipitation is 960 mm. The size of the basins of the Crete region and the values of the annual precipitation constitute a rather homogenous set for the application of the proposed SN–SPI.

The SN–SPI behaves in a similar way to the SPI; therefore, the interpretation of the results is much the same since the thresholds used are the same as SPI. However, a significant difference is while the SPI is temporally comparable, the SN–SPI is spatio-temporally comparable among different areas with different mean total annual precipitation (Fig. 12). Intense fluctuations in the short-term plots (12- and 24-month time scales) once again reveal the seasonality component of precipitation; obviously, the difference between the indices appears mainly at the peaks of the time series. In 12-month SPI and SN–SPI plots, one can observe the normalization taking place during extremely dry and wet conditions (1990: Tavronitis, 1978–2003: Geropotamos, and 1980–1987–2003: Petras). Passing along the next time scale (24 months) the difference among the indices becomes clearer, as the SN–SPI presents Tavronitis as less dry during 1989–91, Geropotamos as less wet during 2003, and Petras as less wet during 1979–82, 1988, and 2003. Referring to the long-term scale (48 months), we deduce that in the case of Tavronitis (wet watershed) normalization takes place during dry conditions (1987–95); whereas in Geropotamos and Petras (dry watersheds) normalization is obvious during wet conditions (1981–2003 and 1981–83–2003, respectively). Although the aforementioned watersheds are considered wet during those particular periods of time, the SN–SPI smooths the extreme conditions and makes it possible to compare these watersheds in relation to the rest by taking into account the spatial character of precipitation.

More specifically, regarding two different watersheds and taking into consideration two occasional time periods and the respective mean precipitation of the previous 48 months (given that SPI and SN–SPI of the 48 previous months are calculated for each year) SPI and SN–SPI differentiation is shown (Table 5). It is important to notice that annual precipitation height for Tavronitis is twice the height of the Petras watershed. Despite this fact, as Fig. 13 illustrates in general, SPI time series show higher values for Petras than Tavronitis. Short-term SPI and SN–SPI variations do not reach to a point so as to provide for a clear comparison between these indices, except the aforementioned seasonality. Nevertheless, it can be said that 12-month SPI values show that Tavronitis experiences several moderate droughts until 1988 and an extreme drought during 1990, whereas Petras has moderately wet conditions until 1985 and extremely wet conditions during 1987 and 2003. SN–SPI values present Tavronitis with mildly dry conditions during the aforementioned periods and Petras with mildly wet conditions. In 24-month plots, the same occurs with fewer fluctuations. The long-term plot of 48-month SPI and SN–SPI clearly provides the aforementioned difference, as the normalization occurs for Petras during 1978–89 (wet conditions) and for Tavronitis during 1987–95 (dry conditions). Stated in the simplest terms, the same SPI value occurs for different precipitation levels and, therefore, these watersheds cannot be compared. On the other hand, different SN–SPI values occur for different precipitation levels. On January 1993, Tavronitis with SPI = −1.72 is considered to be severely dry, while with SN–SPI = −0.83 is mildly dry for all of the 130 watersheds. As one reaches the zero point, the normalization is less obvious; a fact that is shown by the results on January 1996, where mild conditions of drought occur for both watersheds and the difference between SPI and SN–SPI narrows.

The SN–SPI seems to be more appropriate for geographical comparison of neighboring regions with precipitation diversity. Generally speaking, it appears that the SN–SPI values are reasonably comparable in their local significance both in space and time.

Furthermore, the SN–SPI was calculated for 12-, 24-, and 48-month time scales for the generation of the respective extremity maps in terms of time percentage (30 years) (Fig. 14). According to the SN–SPI classification scale, the watersheds were classified in terms of extremely dry conditions as a time percentage, which is defined as the temporal part of extreme dryness out of the 30-year period under study. As far as 12-month time scale is concerned, 3% of the total area of the central and eastern part of Crete experiences extreme dryness the majority of the time (3%–3.9%). A time scale of 24 months results in extreme drought conditions for 16% of the total area in the central and eastern parts 5%–5.9% of the time; while for a 48-month time scale, noteworthy extreme drought characterizes 25% of the whole island 6%–7% of the time. In contrast, Fig. 15 displays the extremity map of drought conditions as a time percentage concerning the SPI, where 96%, 95%, and 80% of the area suffers from extreme drought 6%–7% of the time according to 12-, 24-, and 48-month SPI, respectively. On the basis of this evidence, it appears that these high percentages do not include the objectiveness and the relativity between the watersheds that the SN–SPI takes into account. Furthermore, using SN–SPI the extreme drought area percentage increases when time scale increases, yet with SPI this percentage decreases.

It is crucial to point out that during the period 1991–95, the whole area of Crete has been under drought conditions and reached its peak the years 1992–93, during which the conditions were extremely dry (48-month SN–SPI), while the rest of the time scales report extreme drought the year 1990 (Fig. 16).

A drought scenario was formulated for representing the possible conditions of 2010–2100 based on the climate change predictions of the customized set of 10 regional climate models (future precipitation values) and the trend analysis of SN–SPI data (Fig. 17a). The 12-month SN–SPI plot shows extreme drought peaks—especially after 2050—that seem to decrease in 24-month SN–SPI and are smoothed finally in 48-month SN–SPI. As for the first 30-year period of 48-month SN–SPI, a slightly dry period of 4 years (2037–38) will likely appear as a separate drought spell unless one recognizes that this event constitutes only a slight and probably beneficial interruption in a fairly long period of unusually wet weather. There may be two dry periods after 2040 (2047–60, 2065–70) and two additional dry periods after 2070 with an interval of a 2-year wet period. It seems reasonable to assume that the period 2070–2100 signs the beginning of drought that closely follows the onset of an extended period of unusually mildly dry weather. The trend analysis of 12-, 24-, and 48-month SPI data (Fig. 17b) demonstrates similar results to the SN–SPI for 12- and 24-month time scales; but at a longer-term scale of 48 months reveals differences in intensities and event duration.

More specifically, the long-term means (48-month SN–SPI time scale) that more than half of the total area of Crete is about to experience drought conditions during 28%, 69%, and 97% of the 2010–40, 2040–70, and 2070–2100 time slices, respectively (Fig. 18). The SN–SPI happens also indicates that it is possible that 52%, 33%, and 25% of the area of Crete will suffer from extremely dry conditions during the 90-year future period according in 12-, 24-, and 48-month time frames. It is then obvious that, moving to a longer-term time scale, the future percentage of the extreme drought area decreases.

It is to be noted that the results of the climate models produce a statistical sense to the frequency of drought events; hence it the specific conditions for specific periods in the area under study cannot be concluded with certainty.

5. Conclusions

The SPI and SN–SPI indices were successfully applied in the island of Crete for drought assessment for the period 1973–2004 using three time scales (12, 24, and 48 months). The SPI analysis confirmed that Crete is facing drought problems at the southern and eastern parts. The proposed methodology is easily applied and interpreted the precipitation deficit; thus, it could become a practical tool for the assessment of regional drought events. Accordingly, the SPI is a statistically coherent index for sensitively measuring drought.

The SN–SPI is a variant of the common tool of drought assessment SPI, but represents a more suitable means of comparing drought conditions between neighboring areas of differing precipitation heights because the SN–SPI expands the meaning of the temporal character of drought to its spatial relativity. It appears that these new drought index values are reasonably comparable in their local significance both in space and time. Needless to say, the SN–SPI is spatiotemporally comparable amongst different neighboring regions with different mean total annual precipitation.

Results based on both indices showed that, concerning the 1973–2004 period, the duration of the extreme drought periods remain the same (reaching 7% of time) but the intensities based on SN–SPI are lower. The area covered by extreme droughts is 3% (12 months), 16% (24 months), and 25% (48 months), and 96% (12 months), 95% (24 months), and 80% (48 months) based on the SN–SPI and SPI, respectively.

The future scenario for 2010–2100, formulated by the results of a set of regional climate models, indicates a gradual decrease of precipitation and therefore an extended period of mild drought, which will negatively affect the status of the ecosystems and human environments and may lead to intense water scarcity problems. Concerning the longest time scale (48 months), more than half of the total area of Crete is about to experience drought conditions during 28%, 69%, and 97% of the 2010–40, 2040–70, and 2070–2100 periods, respectively, while extremely dry conditions will cover 52%, 33%, and 25% of the area for the 90-year period according to 12-, 24-, and 48-month SN–SPI, respectively.

During the last decades, the island of Crete has faced a large number of droughts and during the summer of 2008 there were severe water shortages in Mediterranean countries and drastic measures were undertaken to rectify this situation. For example, in the month of July 2008, the water that had to be delivered to the islands of Greece via tankers was 10% more than 2007 at a cost of 11 million euros, and the reservoirs were standing at the lowest level. Moreover, large infrastructures are already constructed in the island of Crete in order to transfer water among neighboring watersheds, mostly for irrigation purposes. Appreciable economic benefit for municipalities within basins with surplus water resources could emerge from transferring excess freshwater amounts to nearby arid regions. This information—the relative comparison of watershed-level water resources conditions—can be provided by the SN–SPI.

Finally, the results signal an urgent need for the development of strategic water management and preparedness plans in all drought-prone areas in order to help mitigate most of the resulting adverse effects. Shorter rainy periods could seriously affect water resources, with wide-ranging consequences for local human societies and ecosystems. The impact of these precipitation changes at the watershed level is required in order to develop strategies in long-term water supply and demand, and thus to attain sustainable water resources management. The policy on droughts is included in the EU Water Framework Directive, and provides a specific framework to adopt in assessing the impact of climate change on water resources and, hence, develop drought management plans that are constantly reviewed. These plans should be updated with the new results presented herein, including the procedure for data analysis and interpretation.

Acknowledgments

The work presented in this paper was supported by the Water and Global Change (WATCH) project. WATCH is an Integrated Project Funded by the European Commission under the Sixth Framework Programme, Global Change and Ecosystems Thematic Priority Area (Contract 036946). The WATCH project started 2 January 2007 and will continue for 4 years.

The ENSEMBLES data used in this work were funded by the EU FP6 Integrated Project ENSEMBLES (Contract 505539), whose support is gratefully acknowledged.

REFERENCES

  • Ahrens, B., 2006: Distance in spatial interpolation of daily rain gauge data. Hydrol. Earth Syst. Sci., 10 , 197208.

  • American Meteorological Society, 2004: AMS statement on meteorological drought. Bull. Amer. Meteor. Soc., 85 , 771773.

  • Ashraf, M., , Loftis J. C. , , and Hubbard K. G. , 1997: Application of geostatistics to evaluate partial weather station networks. Agric. For. Meteor., 84 , 255271.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bacanli, U. G., , Firat M. , , and Dikbas F. , 2009: Adaptive Neuro-Fuzzy Inference System for drought forecasting. Stochastic Environ. Res. Risk Assess., 23 , 11431154.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bonaccorso, B., , Bordi I. , , Cancelliere A. , , Rossi G. , , and Sutera A. , 2003: Spatial variability of drought: An analysis of the SPI in Sicily. Water Resour. Manage., 17 , 273296.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chang, C. L., , Lo S. L. , , and Yu S. L. , 2005: Applying fuzzy theory and genetic algorithm to interpolate precipitation. J. Hydrol., 314 , 92104.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chartzoulakis, K., , and Psarras G. , 2005: Global change effects on crop photosynthesis and production in Mediterranean: The case of Crete, Greece. Agric. Ecosyst. Environ., 106 , 147157.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chartzoulakis, K., , Paranychianakis N. V. , , and Angelakis A. N. , 2001: Water resources management in the island of Crete, Greece, with emphasis on the agricultural use. Water Policy, 3 , 193205.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cheng, C.-H., 1998: A new approach for ranking fuzzy numbers by distance method. Fuzzy Sets Syst., 95 , 307317.

  • Christensen, J. H., , and Christensen O. B. , 2007: A summary of the PRUDENCE model projections of changes in European climate by the end of this century. Climatic Change, 81 , 730. doi:10.1007/s10584-006-9210-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Christensen, J. H., , Boberg F. , , Christensen O. B. , , and Lucas-Picher P. , 2008: On the need for bias correction of regional climate change projections of temperature and precipitation. Geophys. Res. Lett., 35 , L20709. doi:10.1029/2008GL035694.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Christensen, J. H., , Kjellström E. , , Giorgi F. , , Lenderink G. , , and Rummukainen M. , 2010: Weight assignment in regional climate models. Climate Res., 44 , 179194.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Christensen, O. B., , Drews M. , , Christensen J. H. , , Dethloff K. , , Ketelsen K. , , Hebestadt I. , , and Rinke A. , 2006: The HIRHAM Regional Climate Model Version 5 (β). Tech Rep 06-17. ISSN 1399-1388. DMI, Copenhagen.

    • Search Google Scholar
    • Export Citation
  • Collins, M., , Booth B. B. B. , , Bhaskaran B. , , Harris G. R. , , Murphy J. M. , , Sexton D. H. M. , , and Webb M. J. , 2009: Climate model errors, feedbacks and forcings: A comparison of perturbed physics and multi-model ensembles. Climate Dyn., 27 , 127147. doi:10.1007/ s00382-010-0808-0.

    • Search Google Scholar
    • Export Citation
  • Dirks, K. N., , Hay J. E. , , Stow C. D. , , and Harris D. , 1998: High-resolution studies of rainfall on Norfolk Island. Part II: Interpolation of rainfall data. J. Hydrol., 208 , 187193.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • EUR-Lex, cited. 2000: Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for community action in the field of water policy. [Available online at http://ec.europa.eu/environment/water/water-framework/index_en.html].

    • Search Google Scholar
    • Export Citation
  • Founda, D., , and Giannakopoulos C. , 2009: The exceptionally hot summer of 2007 in Athens, Greece—A typical summer in the future climate? Global Planet. Change, 67 , 227236. doi:10.1016/j.gloplacha.2009.03.013.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gaiser, T., and Coauthors, 2008: Development of a regional model for integrated management of water resources at the basin scale. Phys. Chem. Earth, 33 , 175182. doi:10.1016/j.pce.2007.04.018.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Giorgi, F., , and Mearns L. O. , 1999: Introduction to special section: Regional climate modelling revisited. J. Geophys. Res., 104 , 63356352.

  • Hayes, M. J., , Svoboda M. D. , , Wilhite D. A. , , and Vanyarkho O. V. , 1999: Monitoring the 1996 drought using the standardized precipitation index. Bull. Amer. Meteor. Soc., 80 , 429438.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ines, A. V. M., , and Hansen J. W. , 2006: Bias correction of daily GCM rainfall for crop simulation studies. Agric. For. Meteor., 138 , 4453. doi:10.1016/j.agrformet.2006.03.009.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jacob, D., 2001: A note to the simulation of the annual and inter-annual variability of the water budget over the Baltic Sea drainage basin. Meteor. Atmos. Phys., 77 , 6173.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jacob, D., and Coauthors, 2007: An inter-comparison of regional climate models for Europe: Design of the experiments and model performance. Climatic Change, 81 , (Suppl.). 3152.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jacob, D., , Christensen O. B. , , Doblas-Reyes F. J. , , Goodess C. , , Tank A. K. , , Lorenz P. , , and Roeckner E. , 2008: Information on observations, global and regional modelling data availability and statistical downscaling. ENSEMBLES Tech. Rep. 4, 10 pp. [Available online at http://ensembles-eu.metoffice.com/tech_reports/ETR_4_vn1.pdf].

    • Search Google Scholar
    • Export Citation
  • Jaeger, E. B., , Anders I. , , Lüthi D. , , Rockel B. , , Schär C. , , and Seneviratne S. , 2008: Analysis of ERA40-driven CLM simulations for Europe. Meteor. Z., 17 , 349367.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kjellström, E., , Bärring L. , , Gollvik S. , , Hansson U. , , Jones C. , , Samuelsson P. , , Rummukainen M. , , Ullerstig A. , , Willén U. , , and Wyser K. , 2005: A 140-year simulation of European climate with the new version of the Rossby Centre regional atmospheric climate model (RCA3). SMHI Reports Meteorology and Climatology, 108 .Swedish Meteorological and Hydrological Institute, Norrköping, Sweden, 54pp.

    • Search Google Scholar
    • Export Citation
  • Kidson, J. W., , and Thompson C. S. , 1998: A comparison of statistical and model-based downscaling techniques for estimating local climate variations. J. Climate, 11 , 735753.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Komuscu, A. U., 1999: Using the SPI to analyze spatial and temporal patterns of drought in Turkey. Drought Network News, 11 , 713.

  • Koutroulis, A. G., , and Tsanis I. K. , 2010: A method for estimating flash flood peak discharge in a poorly gauged basin: Case study for the 13–14 January 1994 flood, Giofiros basin, Crete, Greece. J. Hydrol., 385 , 150164. doi:10.1016/j.jhydrol.2010.02.012.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Larsén, X. G., , Mann J. , , Göttel H. , , and Jacob D. , 2008: Wind climate and extreme winds from the regional climate model REMO. Proc. European Wind Energy Conf. and Exhibition, Brussels, Belgium, EWEA, 58–62.

    • Search Google Scholar
    • Export Citation
  • Law, A. M., , and Kelton W. D. , 1982: Simulation Modeling and Analysis. McGraw-Hill, 400 pp.

  • Loukas, A., , and Vasiliades L. , 2004: Probabilistic analysis of drought spatiotemporal characteristics in Thessaly region, Greece. Nat. Hazards Earth Syst. Sci., 4 , 719731.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Manios, T., , and Tsanis I. K. , 2006: Evaluating water resources availability and wastewater reuse importance in the water resources management of small Mediterranean municipal districts. Resour. Conserv. Recycl., 47 , 245259.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • McKee, T. B., , Doesken N. J. , , and Kleist J. , 1993: The relationship of drought frequency and duration to time scales. Proc. Eighth Conf. on Applied Climatology, Anaheim, CA, Amer. Meteor. Soc., 179–184.

    • Search Google Scholar
    • Export Citation
  • Meyer, S. J., , Hubbard K. G. , , and Wilhite D. A. , 1993: A crop-specific drought index for corn: I. Model development and validation. Agron. J., 85 , 388395.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Murphy, J., 1999: An evaluation of statistical and dynamical techniques for downscaling local climate. J. Climate, 12 , 22562284.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mylopoulos, Y. A., , and Kolokytha E. G. , 2008: Integrated water management in shared water resources: The EU Water Framework Directive implementation in Greece. Phys. Chem. Earth, 33 , 347353. doi:10.1016/j.pce.2008.02.031.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Naoum, S., , and Tsanis I. K. , 2004: Orographic precipitation modeling with multiple linear regression. J. Hydrol. Eng., 9 , 79102.

  • Palmer, D. J., , Höck B. K. , , Kimberley M. O. , , Watt M. S. , , Lowe D. J. , , and Payn T. W. , 2009: Comparison of spatial prediction techniques for developing Pinus radiata productivity surfaces across New Zealand. For. Ecol. Manage., 258 , 20462055. doi:10.1016/j.foreco.2009.07.057.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Palmer, W., 1965: Meteorological drought. U.S. Weather Bureau Research Paper 45, 58 pp.

  • Rossi, G., 2000: Drought mitigation measures: A comprehensive framework. Drought and Drought Mitigation in Europe, J. Voght and F. Somma, Eds., Kluwer Academic Publishers, 233–246.

    • Search Google Scholar
    • Export Citation
  • Rossi, G., , Benedini M. , , Tsakiris G. , , and Giakoumakis S. , 1992: On regional drought estimation and analysis. Water Resour. Manage., 6 , 249277.

  • Shepard, D., 1968: A two-dimensional interpolation function for irregularly spaced data. Proc. 23rd ACM National Conf., Princeton, NJ, Association of Computing Machinery, 517–524, doi:10.1145/800186.810616.

    • Search Google Scholar
    • Export Citation
  • Siska, P. P., , and Hung I.-K. , 2001: Assessment of kriging accuracy in the GIS environment. Proc. 21st Annual ESRI Int. User Conf., San Diego, CA, Environmental Systems Research Institute, Paper 280.

    • Search Google Scholar
    • Export Citation
  • Solomon, S., , Qin D. , , Manning M. , , Marquis M. , , Averyt K. , , Tignor M. M. B. , , Miller H. L. Jr., , and Chen Z. , Eds. 2007: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.

    • Search Google Scholar
    • Export Citation
  • Sönmez, F. K., , Kömüscü , , Erkan A. , , and Turgu E. , 2005: An analysis of spatial and temporal dimension of drought vulnerability in Turkey using the standardized precipitation index. Nat. Hazards, 35 , 243264. doi:10.1007/s11069-004-5704-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tabios, G. Q., , and Salas J. D. , 1985: A comparative analysis of techniques for spatial interpolation of precipitation. J. Amer. Water Resour. Assoc., 21 , 365380.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tannehill, I. R., 1947: Drought—Its Causes and Effects. Princeton University Press, 597 pp.

  • Teegavarapu, R. S. V., , Tufail M. , , and Ormsbee L. , 2009: Optimal functional forms for estimation of missing precipitation data. J. Hydrol., 374 , 106115.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tsakiris, G., , and Vangelis H. , 2004: Towards a drought watch system based on spatial SPI. Water Resour. Manage., 18 , 112.

  • Tsakiris, G., , Pangalou D. , , Tigkas D. , , and Vangelis H. , 2007: Assessing the areal extent of drought. Proc. EWRA Int. Symp. on Water Resource Management: New Approaches and Technologies, Chania, Greece, European Water Resources Association, 59–66.

    • Search Google Scholar
    • Export Citation
  • Tsanis, I. K., , and Naoum S. , 2003: The effect of spatially distributed meteorological parameters on irrigation water demand assessment. Adv. Water Res., 26 , 311324.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Van der Linden, P., , and Mitchell J. F. B. , Eds. 2009: ENSEMBLES: Climate change and its impacts: Summary of research and results from the ENSEMBLES project. Met Office Hadley Centre, 160 pp. [Available from Met Office Hadley Centre, FitzRoy Road, Exeter EX1 3PB, United Kingdom].

    • Search Google Scholar
    • Export Citation
  • Van Meijgaard, E., , van Ulft L. H. , , van de Berg W. J. , , Bosveld F. C. , , van den Hurk B. J. J. M. , , Lenderink G. , , and Siebesma A. P. , 2008: The KNMI regional atmospheric climate model RACMO, version 2.1. KNMI Publ. TR-302, KNMI, De Bilt, pp. 50.

    • Search Google Scholar
    • Export Citation
  • Ware, C., , Knight W. , , and Wells D. , 1991: Memory intensive statistical algorithms for multibeam bathymetric data. Comput. Geosci., 17 , 985993.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wei, T. C., , and McGuinness J. L. , 1973: Reciprocal Distance Squared Method: A computer technique for estimating area precipitation. U.S. Agricultural Research Service Tech. Rep. ARS-Nc-8., 29 pp.

    • Search Google Scholar
    • Export Citation
  • Wilby, R. L., , Hay L. E. , , Gutowski W. J. Jr., , Arritt R. W. , , Takle E. S. , , Pan Z. , , Leavesley G. H. , , and Clark M. P. , 2000: Hydrological responses to dynamically and statistically downscaled climate model output. Geophys. Res. Lett., 27 , 11991202.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wilhite, D. A., 2000: Drought as a natural hazard: Concepts and definitions. Drought: A Global Assessment, Natural Hazards and Disasters Series, Vol. 2, Routledge Publishers, 3–18.

    • Search Google Scholar
    • Export Citation
  • Wood, A. W., , Maurer E. P. , , Kumar A. , , and Lettenmaier D. P. , 2002: Long-range experimental hydrologic forecasting for the eastern United States. J. Geophys. Res., 107 , 4429. doi:10.1029/2001JD000659.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wu, H., , Svoboda M. D. , , Hayes M. J. , , Wilhite D. A. , , and Wen F. , 2007: Appropriate application of the Standardized Precipitation Index in arid locations and dry seasons. Int. J. Climatol., 27 , 6579.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yamoah, C. F., , Walters D. T. , , Shapiro C. A. , , Francis C. A. , , and Hayes M. J. , 2000: Standardized precipitation index and nitrogen rate effects on crop yields and risk distribution in maize. Agric. Ecosyst. Environ., 80 , 113120.

    • Crossref
    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Hypothetical region consisting of A, B, and C watersheds.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 2.
Fig. 2.

Monthly precipitation time series for the three hypothetical watersheds of Fig. 1.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 3.
Fig. 3.

Monthly SPI 48 (48-month time scale) time series for the 3 hypothetical watersheds.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 4.
Fig. 4.

Area of study, location of watersheds of Tavronitis, Geropotamos, and Petras, and ENSEMBLES grid over Crete Island.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 5.
Fig. 5.

Elevation map of Crete Island.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 6.
Fig. 6.

Mean annual precipitation time series for Tavronitis, Geropotamos, and Petras watersheds for the period 1973–2004.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 7.
Fig. 7.

Average annual precipitation at watershed level for the 1973–2004 period and location of precipitation stations over Crete.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 8.
Fig. 8.

Monthly SN–SPI 48 (48-month time scale) time series for the 3 hypothetical watersheds.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 9.
Fig. 9.

The SN–SPI sensitivity to the precipitation variability of dry basin A for basin (a) A, (b) B, and (c) C.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 10.
Fig. 10.

As in Fig. 9, but for wet basin C.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 11.
Fig. 11.

The (top) 12-, (middle) 24-, and (bottom) 48-month time scale SPI for (a) Tavronitis, (b) Geropotamos, and (c) Petras for 1973–2004.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 12.
Fig. 12.

As in fig. 11, but for the SPI and SN–SPI time series.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 13.
Fig. 13.

As in Fig. 12, but for SPI and SN–SPI time series and only Tavronitis and Petras.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 14.
Fig. 14.

Drought extremity map in terms of time percentage for 1973–2004 for 12-, 24-, and 48-month SN–SPI.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 15.
Fig. 15.

As in Fig. 14, but for the SPI.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 16.
Fig. 16.

Crete percentage area that experiences drought conditions (−1 < SN–SPI ≤ 0; −1,5 < SN–SPI ≤ −1; −2 < SN–SPI ≤ −1,5; and SN–SPI ≤ −2—mildly dry, moderately dry, severely dry, and extremely dry, respectively) for 1973–2004 for (top to bottom) the 12-, 24-, and 48-month SN–SPI.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 17.
Fig. 17.

The 12-, 24-, and 48-month (a) SN–SPI and (b) SPI for Tavronitis, Geropotamos, and Petras for 2010–2100.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Fig. 18.
Fig. 18.

As in Fig. 16, but for 2010–2100.

Citation: Journal of Hydrometeorology 12, 2; 10.1175/2010JHM1252.1

Table 1.

Annual inputs and outputs in the hydrological budget of the island of Crete during normal, humid, and dry years (×109 m3 of water). Source: Chartzoulakis et al. (2001). Values in parentheses are the percentage losses compared to annual input (precipitation).

Table 1.
Table 2.

Annual precipitation in mm for three watersheds: Tavronitis, Geropotamos, and Petras.

Table 2.
Table 3.

Thresholds of the SPI for drought characterization.

Table 3.
Table 4.

List of ENSEMBLES RCMs, corresponding driving GCMs, and final weighted aggregated values based on F1 and F2 metrics.

Table 4.
Table 5.

Example of differences between SPI and SN–SPI values.

Table 5.
Save