• Alpert, P., , B. U. Neeman, , and Y. Shay-El, 1990: Climatological analysis of Mediterranean cyclones using ECMWF data. Tellus, 42 , 6577.

    • Search Google Scholar
    • Export Citation
  • Alpert, P., , M. Tzidulko, , and D. Izigsohn, 1999: A shallow short-lived meso-beta cyclone over the gulf of Antalya, eastern Mediterranean. Tellus, 51A , 249262.

    • Search Google Scholar
    • Export Citation
  • Buzzi, A., , N. Tartaglione, , and P. Malguzzi, 1998: Numerical simulations of the 1994 Piedmont flood: Role of orography and moist processes. Mon. Wea. Rev, 126 , 23692383.

    • Search Google Scholar
    • Export Citation
  • Caracena, F., , R. A. Maddox, , L. R. Hoxit, , and C. F. Chappell, 1979: Mesoanalysis of the Big Thompson storm. Mon. Wea. Rev, 107 , 117.

  • Christian, H. J., and Coauthors, 1999: The lightning imaging sensor. Proc. 11th Int. Conf. on Atmospheric Electricity, Huntsville, AL, ICAE, 746–749.

  • Cummins, K. L., , M. J. Murphy, , E. A. Bardo, , W. L. Hiscox, , R. B. Pyle, , and A. E. Pifer, 1998: A combined TOA/MDF technology upgrade of the U.S. National Lightning Detection Network. J. Geophys. Res, 103 , D8. 90359044.

    • Search Google Scholar
    • Export Citation
  • Dietrich, S., , C. Adamo, , A. Mugnai, , and R. Bechini, 2000: A physical profile-based multisensor precipitation retrieval. Proc. 2000 EUMETSAT Meteorological Satellite Data Users' Conf., Bologna, Italy, EUMETSAT, Vol. EUM P29, 443–449.

  • Ducrocq, V., , D. Ricard, , J-P. Lafore, , and F. Orain, 2002: Storm-scale numerical rainfall prediction for five precipitating events over France: On the importance of the initial humidity field. Wea. Forecasting, 17 , 12361256.

    • Search Google Scholar
    • Export Citation
  • Dudhia, J., 1993: A non-hydrostatic version of the Penn State/NCAR Mesoscale Model: Validation tests and simulation of an Atlantic cyclone and cold front. Mon. Wea. Rev, 121 , 14931513.

    • Search Google Scholar
    • Export Citation
  • Ferretti, R., , S. Low-Nam, , and R. Rotunno, 2000: Numerical simulations of the Piedmont flood of 4-6 November 1994. Tellus, 52 , 162180.

    • Search Google Scholar
    • Export Citation
  • Grossman, R. L., , and D. R. Durran, 1984: Interaction of low-level flow with the western Ghat Mountains and offshore convection in the summer monsoon. Mon. Wea. Rev, 112 , 652672.

    • Search Google Scholar
    • Export Citation
  • Holt, M. A., , P. J. Hardaker, , and G. P. McLelland, 2001: A lightning climatology for Europe and the UK, 1990-99. Weather, 56 , 290296.

  • Hong, S-Y., , and H-L. Pan, 1996: Nonlocal boundary layer vertical diffusion in a medium-range forecast model. Mon. Wea. Rev, 124 , 23222339.

    • Search Google Scholar
    • Export Citation
  • Jansà, A., , A. Genovés, , and J. A. García-Moya, 2000: Western Mediterranean cyclones and heavy rain. Part 1: Numerical experiment concerning the Piedmont flood case. Meteor. Appl, 7 , 323333.

    • Search Google Scholar
    • Export Citation
  • Kain, J. S., , and J. M. Fritsch, 1993: Convective parameterization for mesoscale models: The Kain-Fritsch scheme. The Representation of Cumulus in Numerical Models, Meteor. Monogr., No. 46, Amer. Meteor. Soc., 165–177.

  • Kotroni, V., , and K. Lagouvardos, 2001: Precipitation forecast skill of different convective parameterization and microphysical schemes: Application for the cold season over Greece. Geophys. Res. Lett, 28 , 19771980.

    • Search Google Scholar
    • Export Citation
  • Kotroni, V., , K. Lagouvardos, , G. Kallos, , and D. Ziakopoulos, 1999: Severe flooding over central and southern Greece associated with pre-cold frontal orographic lifting. Quart. J. Roy. Meteor. Soc, 125 , 967991.

    • Search Google Scholar
    • Export Citation
  • Krichak, S. O., , P. Alpert, , and M. Dayan, 2004: Role of atmospheric processes associated with Hurricane Olga in the December 2001 floods in Israel. J. Hydrometeor, 5 , 12591270.

    • Search Google Scholar
    • Export Citation
  • Kummerow, C. W., , W. S. Olson, , and L. Giglio, 1996: A simplified scheme for obtaining precipitation and vertical hydrometeors profiles from passive microwave sensors. IEEE Trans. Geosci. Remote Sens, 34 , 12131232.

    • Search Google Scholar
    • Export Citation
  • Kummerow, C., , W. Barnes, , T. Kozu, , J. Shiue, , and J. Simpson, 1998: The Tropical Rainfall Measuring Mission (TRMM) sensor package. J. Atmos. Oceanic Technol, 15 , 809817.

    • Search Google Scholar
    • Export Citation
  • Kurz, M., , and A. D. Fontana, 2004: A case of cyclogenesis over the western Mediterranean Sea with extraordinary convective activity. Meteor. Appl, 11 , 97113.

    • Search Google Scholar
    • Export Citation
  • Lagouvardos, K., , and V. Kotroni, 1999: Aspects of the 25-26 March 1998 storm over Greece. Weather, 54 , 154160.

  • Lagouvardos, K., , and V. Kotroni, 2000: Use of METEOSAT water-vapour images for the diagnosis of a vigorous stratospheric intrusion over Central Mediterranean. Meteor. Appl, 7 , 205210.

    • Search Google Scholar
    • Export Citation
  • Lagouvardos, K., , and V. Kotroni, 2005: Improvement of high resolution weather forecasts through humidity adjustment, based on satellite data. Quart. J. Roy. Meteor. Soc, 131 , 26952712.

    • Search Google Scholar
    • Export Citation
  • Lagouvardos, K., , V. Kotroni, , S. Dobricic, , S. Nickovic, , and G. Kallos, 1996: On the storm of 21-22 October 1994 over Greece: Observations and model results. J. Geophys. Res, 101 , D21. 2621726226.

    • Search Google Scholar
    • Export Citation
  • Lagouvardos, K., , V. Kotroni, , S. Nickovic, , D. Jovic, , and G. Kallos, 1999: Observations and model simulations of a winter sub-synoptic vortex over the central Mediterranean. Meteor. Appl, 6 , 371383.

    • Search Google Scholar
    • Export Citation
  • Lee, A. C. L., 1986: An experimental study of the remote location of lightning flashes using VLF arrival time difference technique. Quart. J. Roy. Meteor. Soc, 112 , 203229.

    • Search Google Scholar
    • Export Citation
  • Maheras, P., , H. A. Flocas, , C. Anagnostopoulou, , and I. Patrikas, 2002: On the vertical structure of composite surface cyclones in the Mediterranean region. Theor. Appl. Climatol, 71 , 3–4. 199217.

    • Search Google Scholar
    • Export Citation
  • Negri, A. J., , R. F. Adler, , and P. J. Wetzel, 1984: Rain estimation from satellite: An examination of the Griffith–Woodley technique. J. Climate Appl. Meteor, 23 , 102116.

    • Search Google Scholar
    • Export Citation
  • Ogura, Y., , T. Asai, , and K. Dohi, 1985: A case study of heavy precipitation event along the Baiu front in northern Kyushu, 23 July 1982: Nagasaki heavy rainfall. J. Meteor. Soc. Japan, 63 , 883900.

    • Search Google Scholar
    • Export Citation
  • Pinori, S., and Coauthors, 2001: MW/IR satellite data for a combined rainfall estimation of Piemonte and Valle d'Aosta 2000 flood event. Proc. Third EGS Plinius Conf. on Mediterranean Storms, Baja Sardinia, Italy, European Geophysical Society, 161–170.

  • Porcù, F., , F. Prodi, , S. Dietrich, , A. Mugnai, , and R. Bechini, 2000: Multisensor estimation of severe rainfall events. Proc. 2000 Eumetsat Meteorological Satellite Data Users' Conf., Darmstadt, Germany, EUMETSAT, Vol. EUM P29, 371–378.

  • Porcù, F., , C. Caracciolo, , and F. Prodi, 2003: Cloud systems leading to flood events in Europe: Overview and classification. Meteor. Appl, 10 , 217227.

    • Search Google Scholar
    • Export Citation
  • Schultz, P., 1995: An explicit cloud physics parameterization for operational numerical weather prediction. Mon. Wea. Rev, 123 , 33313343.

    • Search Google Scholar
    • Export Citation
  • Todd, M. C., , C. Kidd, , D. Kniverton, , and T. J. Bellerby, 2001: A combined satellite infrared and passive microwave technique for estimation of small-scale rainfall. J. Atmos. Oceanic Technol, 18 , 742755.

    • Search Google Scholar
    • Export Citation
  • Trigo, I. F., , G. R. Bigg, , and T. D. Davies, 2002: Climatology of cyclogenesis mechanisms in the Mediterranean. Mon. Wea. Rev, 130 , 549569.

    • Search Google Scholar
    • Export Citation
  • Tripoli, G. J., , G. Panegrossi, , S. Dietrich, , A. Mugnai, , E. A. Smith, , and F. Siccardi, 2001: A numerical study of the 13-16 October, 2000 Piemonte flood: Comparison of meteorological conditions to two other contrasting fall season flooding events in northern Italy. Proc. Third EGS Plinius Conf. on Mediterranean Storms, Baja Sardinia, Italy, European Geophysical Society, 115–119.

  • Tsidulko, M., , and P. Alpert, 2001: Synergism of upper-level potential vorticity and mountains in Genoa lee cyclogenesis: A numerical study. Meteor. Atmos. Phys, 78 , 261285.

    • Search Google Scholar
    • Export Citation
  • Turk, F. J., , S. Chi-Sann Liou, , R. Qiu, , M. Scofield, , B. Ba, , and A. Gruber, 2000: Capabilities and characteristics of rainfall estimates from geostationary- and geostationary + microwave-based satellite techniques. Preprints. Symp. on Precipitation Extremes: Prediction, Impacts, and Responses, Albuquerque, NM, Amer. Meteor. Soc., 191–194.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    (a) Analysis of mean sea level pressure at 2-hPa intervals from the ECMWF at 0000 UTC 4 Dec. Equivalent potential temperature values (at 850 hPa) greater than 308 K are shaded at 2-K interval. (b) Analysis of the 500-hPa-level geopotential height at 40-m interval. Shading denotes wind speed at 300-hPa level contoured at 10 m s−1 interval (only values exceeding 40 m s−1 are plotted). (c) and (e) As in (a), but at 0000 and 1800 UTC 5 Dec, respectively; (d) and (f) as in (b), but for 0000 and 1800 UTC 5 Dec, respectively.

  • View in gallery

    The 6-h-accumulated rainfall (mm) from the surface network reports, ending at 1200 and 1800 UTC 5 Dec 2002 and 0000 UTC 6 Dec 2002. Respective values are separated by a slash. Antalya is denoted by an asterisk.

  • View in gallery

    Geolocations of the CG flashes in 3-h intervals as recorded by the ATD system for the period from 0600 UTC 5 Dec 2002 to 0600 UTC 6 Dec 2002.

  • View in gallery

    (a) Time series of the CG flash rate and (b) maximum of flash density, calculated in 5-min intervals for the period from 0600 UTC 5 Dec 2002 to 0600 UTC 6 Dec 2002. Flash density was computed by 0.2° lat × 0.2° lon.

  • View in gallery

    TMI 85-GHz horizontal polarization images over Antalya region during (a) the first overpass at 1631 UTC 5 Dec and (b) the second overpass at 1808 UTC the same day. Examples of two flashes recorded by LIS at (c) 1807:55 and (d) 1808:27 UTC (see text for details) are shown. (e) The shades of gray coded time series of the maximum of optical magnitude per LIS frame recorded during the flash shown in (c); (f) as in (e), but for the flash shown in (d). Triangle in (f) shows the time of the ATD fix recorded during the same flash. Its location is indicated at the intersection of the horizontal and vertical solid lines in (d).

  • View in gallery

    The 6-h-accumulated rainfall ending at (a) 1800 UTC 5 Dec 2002 and at (b) 0000 UTC 6 Dec 2002 as calculated by the NAW method.

  • View in gallery

    PMW rain-rate estimates (mm h−1) from the (a) SSM/I overpass (1531 UTC), (b) TRMM overpass (1631 UTC), and (c) TRMM overpass (1808 UTC).

  • View in gallery

    The 6-h-accumulated rainfall ending at (a) 1800 UTC 5 Dec 2002 and at (b) 0000 UTC 6 Dec 2002 as calculated by the PMW–IR method.

  • View in gallery

    (a) Mean sea level pressure at 2-hPa intervals and (b) 500-hPa geopotential height at 40-m interval. Shading denotes wind speed at 300-hPa level contoured at 10 m s−1 interval (only values exceeding 40 m s−1 are plotted) valid at 0000 UTC 5 Dec 2002, as predicted by MM5 grid 1.

  • View in gallery

    (a) The 925-hPa equivalent potential temperature (at 2-K intervals; values greater than 308 K are shaded), (b) The 925-hPa wind speed (contoured at 2 m s−1 interval; only values greater than 10 m s−1 are shown) from MM5 grid 2 (CNTRL run), valid at 1200 UTC 5 Dec 2002. (c) As in (a), but at 1800 UTC 5 Dec, (d) as in (b), but at 1800 UTC 5 Dec, (e) as in (a) but at 0000 UTC 6 Dec, and (f) as in (b), but at 0000 UTC 6 Dec.

  • View in gallery

    The 6-h-accumulated precipitation (at 10-mm interval) valid at (a) 1200 UTC 5 Dec 2002, (b) 1800 UTC 5 Dec 2002, and (c) 0000 UTC 6 Dec 2002 as predicted by MM5 grid 2 (CNTRL run). The mountain barrier around Antalya is denoted by a thick line.

  • View in gallery

    The 6-h-accumulated precipitation (at 10-mm interval) valid at (a) 1200 UTC 5 Dec 2002, (b) 1800 UTC 5 Dec 2002, and (c) 0000 UTC 6 Dec 2002 as predicted by MM5 grid 2 (ASSIM run).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 3 3 3
PDF Downloads 2 2 2

The Antalya 5 December 2002 Storm: Observations and Model Analysis

View More View Less
  • 1 Institute for Environmental Research, National Observatory of Athens, Athens, Greece
  • 2 Institute of Atmospheric Sciences and Climate, CNR, Cagliari, Italy
  • 3 Department of Physics, University of Ferrara, Ferrara, Italy
  • 4 Institute of Atmospheric Sciences and Climate, CNR, Cagliari, Italy
  • 5 Turkish State Meteorological Service, Ankara, Turkey
© Get Permissions
Full access

Abstract

In the frame of this work, the storm that occurred on 5 December 2002 in Antalya, located on the southwestern Mediterranean Sea coast of Turkey, is analyzed. More than 230 mm of 24-h-accumulated rainfall have been reported during the event that produced floods in the area. The analysis is based on the results of model simulations with the fifth-generation Pennsylvania State University–NCAR Mesoscale Model (MM5). Observational data provided by the Tropical Rainfall Measuring Mission (TRMM) sensors (including the Lightning Imaging Sensor and TRMM Microwave Imager), Special Sensor Microwave Imager (SSM/I), Meteosat-7, and Met Office Arrival Time Difference (ATD) lightning network are used for both the comparison with the model results and also for the characterization of the storm. The synergetic use of all of this information was crucial for the description of the event. The maximum of precipitation was associated with the warm and moist air masses driven by a low-level jet over the area and impinging over the orographic barriers. The improvement of representation of the humidity field in the model initial conditions, through a simple technique of humidity adjustment based on satellite rainfall estimates, resulted in an improvement of the prediction of the timing and quantity of the precipitation maxima during the event.

Corresponding author address: Dr. V. Kotroni, National Observatory of Athens, Institute of Environmental Research and Sustainable Development, Lofos Koufou, P. Penteli, 15236, Athens, Greece. Email: kotroni@meteo.noa.gr

Abstract

In the frame of this work, the storm that occurred on 5 December 2002 in Antalya, located on the southwestern Mediterranean Sea coast of Turkey, is analyzed. More than 230 mm of 24-h-accumulated rainfall have been reported during the event that produced floods in the area. The analysis is based on the results of model simulations with the fifth-generation Pennsylvania State University–NCAR Mesoscale Model (MM5). Observational data provided by the Tropical Rainfall Measuring Mission (TRMM) sensors (including the Lightning Imaging Sensor and TRMM Microwave Imager), Special Sensor Microwave Imager (SSM/I), Meteosat-7, and Met Office Arrival Time Difference (ATD) lightning network are used for both the comparison with the model results and also for the characterization of the storm. The synergetic use of all of this information was crucial for the description of the event. The maximum of precipitation was associated with the warm and moist air masses driven by a low-level jet over the area and impinging over the orographic barriers. The improvement of representation of the humidity field in the model initial conditions, through a simple technique of humidity adjustment based on satellite rainfall estimates, resulted in an improvement of the prediction of the timing and quantity of the precipitation maxima during the event.

Corresponding author address: Dr. V. Kotroni, National Observatory of Athens, Institute of Environmental Research and Sustainable Development, Lofos Koufou, P. Penteli, 15236, Athens, Greece. Email: kotroni@meteo.noa.gr

Keywords: Lightning

1. Introduction

The Mediterranean Sea is an important cyclogenetic area (Alpert et al. 1990) where the air–sea interaction over the relatively warm Mediterranean waters, but also the complex topography and land water distribution, under favorable conditions lead to either the development or regeneration of cyclones. Many authors have devoted a number of studies to the analysis of Mediterranean cyclones either from a climatological (Alpert et al. 1990; Trigo et al. 2002; Maheras et al. 2002) or a dynamics (Lagouvardos et al. 1996; Lagouvardos and Kotroni 1999, 2000; Kotroni et al. 1999, Jansà et al. 2000; Kurz and Fontana 2004) point of view. The orographic influence and the role of surface fluxes over the Mediterranean waters on the convective activity development and lifetime has been also stressed (Buzzi et al. 1998; Ferretti et al. 2000; Lagouvardos et al. 1999; Tsidulko and Alpert 2001).

Over the Mediterranean region, heavy rainfall events occur throughout the year, but the heaviest events, especially over the central and eastern Mediterranean, mainly occur during autumn and winter, producing significant rain amounts leading to flooding (Porcù et al. 2003). As an example, we mention the extensive flooding in Piedmont (Italy) in November 1994 (Buzzi et al. 1998; Ferretti et al. 2000), in the Po Valley (Italy) in October 2000 (Pinori et al. 2001; Tripoli et al. 2001), in Peloponnissos (Greece) in January 1997 (Kotroni et al. 1999) and Athens (Greece) in October 1994 (Lagouvardos et al. 1999), and over the southern part of Turkey and the Black Sea during the winter of 2001/02.

In the frame of this study the storm that occurred during the period of 4–6 December 2002 in the Antalya region (located in the southwestern Mediterranean coast of Turkey) is analyzed. More than 230 mm of 24-h-accumulated rainfall have been reported during the event that produced floods in the area. The analysis is based on the available observations and the results of model simulations. The fifth-generation Pennsylvania State University–National Center for Atmospheric Research Mesoscale Model (MM5) is used in order to analyze the dynamics of the event. Observational data provided by the Tropical Rainfall Measuring Mission (TRMM) sensors [including the Lightning Imaging Sensor (LIS) and TRMM Microwave Imager (TMI)], Special Sensor Microwave Imager (SSM/I), Meteosat-7, and Met Office Arrival Time Difference (ATD) lightning network system are used for both the comparison with the model results and for the characterization of the storm. The rainfall estimates based on satellite data have been also used for the adjustment of humidity in the initial conditions of the model simulation in order to investigate the role of a better description of humidity in the improvement of the precipitation forecasts.

The paper is structured as follows. A brief description of the synoptic and mesoscale features of the event is given in the following section. Satellite observations and lightning reports are discussed in section 3, and section 4 is devoted to the model description and setup. The simulation results and the comparison with the observations are discussed in section 5. The last section is devoted to the summary of the findings.

2. Synoptic setup

The presentation of the case at this stage is based on the European Centre for Medium-Range Weather Forecasts (ECMWF) analysis at 0.5° resolution. On 0000 UTC 4 December 2002 a surface low of 998 hPa was centered over southern Italy (Fig. 1a) with a cold front spanning from north to south over the Adriatic Sea (not shown). The warm sector of the surface front is characterized by warm and moist air masses that are conveyed toward the southern Adriatic and the Aegean Sea, as depicted by the equivalent potential temperature field at 850 hPa (Fig. 1a). At 500 hPa a cutoff low of 5385 m was located slightly to the west of the surface low with the axis of the trough being oriented northwest–southeast (Fig. 1b). The southern Aegean Sea and the eastern–southeastern part of Turkey were under the influence of the diffluent trough. The upper-level jet, as depicted at the 300-hPa level (Fig. 1b), had a north–south orientation with a cyclonic curvature at the southern flank of the cutoff low evident at the 500-hPa level.

During the following 24 h the low pressure center has been stationary with a central pressure of 1000 hPa over southern Italy (Fig. 1c), while the associated cold front has progressed northeastward, turning cyclonically, and spanned through the Aegean Sea down to the northern coast of Libya (at 0000 UTC 5 December). A warm conveyor belt is evident at the 850-hPa-equivalent potential temperature field, spanning through the eastern Aegean and the western Turkish coasts down to the maritime area west of Cyprus, and is associated with southerly winds of ∼15 m s−1 at this level (not shown). Inspection of the upper-air maps shows that the cutoff low also stayed almost stationary without any deepening while only the axis of the trough slowly turned cyclonically (Fig. 1d). The upper-level jet has also progressed slowly and the area of southern Greece is under its left exit region, a feature that should favor ascending motions in the area.

The maximum of rainfall in the Antalya region has occurred up to 1800 UTC; this will be discussed here following. At that time, the center of the surface low is still over southern Italy, but the frontal discontinuity defined by the temperature and wind gradient has progressed eastward and the area of the southwestern part of Turkey, and especially Antalya, is affected by the advection of a narrow belt of warm and moist air masses (Fig. 1e). This belt of warm and moist air while progressing over the sea impinges the mountains that encircle Antalya from west to east. Thus, an orographic enhancement of rainfall is expected in the area. The importance of orographic lifting over the Gulf of Antalya has been also evidenced by Alpert et al. (1999) in their study of a shallow mesobeta cyclone over the same area. At 500 hPa the trough axis has further turned cyclonically, together with the upper-level jet, which at 1800 UTC 5 December (the time of maximum rainfall over Antalya) was spanning almost zonally over the southeastern Mediterranean, with its left exit region being located over the maritime area south of Antalya (Fig. 1f).

As can be seen in Fig. 2, where the 6-h-accumulated rainfalls ending at 1200, 1800, and 0000 UTC 5–6 December 2002 are given, the event was not accompanied by significant rainfall, except for the region around Antalya Bay. In the area around Antalya, 70 and 208 mm of rain were reported by two stations in 18 h, with a maximum 6-h accumulation of 122 mm at 1800 UTC at the Antalya station. This maximum was associated with the tongue of low-level flow of moist and warm air masses depicted in Fig. 1e.

3. Analysis of satellite and lightning data

The following section focuses on ground-based and spaceborne observations of lightning activity and cloud properties recorded on 5 December when the maximum of rainfall was measured over Antalya.

a. Lightning activity

The lightning activity has been studied based on observations from (a) the Met Office long-range series ATD system and (b) the spaceborne National Aeronautics and Space Administration Lightning Imaging Sensor on board TRMM. The ATD system is designed to determine the locations of cloud-to-ground (CG) flashes by sensing the specific very low frequency radiation emitted during the ground connections (Lee 1986; Holt et al. 2001). Because a CG flash can be composed of many ground connections (also called ATD fixes), an algorithm was developed that combines the ATD fixes in flashes. This algorithm is similar to the one developed and used in the U.S. National Lightning Detection Network (Cummins et al. 1998). Two ATD fixes are part of the same CG flash if the time interval between the two fixes is less than 0.5 s and the distance between the two fixes is less than 10 km. The duration of a CG flash is defined by the time interval between the first and the last ATD fixes of the flash and is assumed to be less than 1 s. The maximum number of fixes per flash is assumed to be less than 15. Based on the list of flashes as deduced from the algorithm, the CG flash rate was determined as well as the CG flash density on a 0.2° × 0.2° mesh over the study area.

Data provided by LIS on board the TRMM satellite (Kummerow et al. 1998; Christian et al. 1999) have been also analyzed. LIS is an optical device designed to sense at 777 nm the light radiated by the lightning flashes (Christian et al. 1999), and it can record both intracloud (IC) and CG flashes. The lightning observations from LIS have been compared with those from ATD during the two TRMM passes over the Antalya region in order to document the ratio of the CG lightning flashes relative to the total (IC and CG) lightning flashes [CG/(IC + CG)].

Figure 3 shows the geolocations of CG flashes recorded by the ATD system over the Antalya region during a 24-h period beginning at 0600 UTC 5 December 2002. During approximately 12 h (from 1300 UTC 5 December to 0100 UTC 6 December) the CG lightning activity was almost stationary over the Antalya region, while later the CG activity moved eastward along the Turkish coastline.

The CG flash rate computed per 5-min period (Fig. 4) showed that the ATD system sensed more than 3300 flashes during the 24-h period, while among them about 2300 flashes were recorded during the 12 h of the stationary phase over the Antalya region. The CG lightning activity was continuous and composed of successive “intense” and “weak” phases (Fig. 4a). The CG flash rate reached a maximum of 37 flashes per 5 min roughly at the end of the stationary period (∼0100 UTC 6 December), while later on the flash rate decreased substantially. The maximum CG flash density in the studied region (Fig. 4b) was higher (with a peak of 12 flashes per 5 min per 0.2° × 0.2°) in the beginning of the period when the activity was concentrated in a small area around Antalya (see upper panel in Fig. 3). Later on, the maximum of the CG density diminishes as the lightning activity becomes more spread around Antalya.

Figures 5a and 5b show the TMI at 85 GHz (horizontal polarization) during the two overpasses on 5 December 2002 (1631 and 1808 UTC, respectively). Colder brightness temperatures at 85 GHz were recorded during the second pass (135 versus 168 K). Analysis of the LIS data showed that nine flashes and a maximum rate of six flashes per minute were recorded over the area of interest during the 90-s LIS time window of the first overpass, while 19 flashes and a maximum flash rate of 11 flashes per minute were recorded during the second overpass. This feature indicates a stronger electrical activity during the second overpass, suggesting deeper convection at that time.

Further, the analysis revealed that two categories of flashes occurred during the Antalya storm. The first category consists of “regular” flashes. They are typically composed of few components (as sensed by LIS) and they last on the average for about 0.5 s. Figures 5c and 5e show an example of such a flash. The second category of flashes is composed of extensive flashes (up to 100 km long) with duration of up to 2 s, as shown in Figs. 5d and 5f. Those extensive flashes also exhibited bright components with a relatively long duration (up to a few hundred milliseconds). LIS reported one flash of the second category in each overpass. The long flash shown in Fig. 5d has produced a single fix to the ground that has been recorded by the ATD system (shown by the intersection of two perpendicular lines in Fig. 5d). The occurrence of the second flash category suggests that the storm exhibited complex and extensive charged regions.

Last, we noted a temporal and spatial consistency in the measurements of the two lightning sensors (ATD and LIS). A flash-by-flash analysis revealed that during the first overpass, 3 out of the 9 LIS flashes were recorded by ATD, while during the second overpass 4 out of the 19 LIS flashes were sensed by ATD. It implies that the ratio of CG/(IC + CG) was 0.33 and 0.21, respectively, during the first and the second overpasses.

b. Satellite rainfall estimates

For the purposes of this study, rainfall estimates have been also produced based on satellite data with the aim to describe the Antalya storm but also to compare with the model simulations and to assimilate these data as described later in section 4. Namely, rainfall estimates were carried out by means of the Negri–Adler–Wetzel (NAW) technique (Negri et al. 1984) as a baseline infrared-only algorithm and a passive microwave (PMW)–infrared technique (Porcù et al. 2000) to take advantage of the three PMW overpasses during this event. Indeed, the most intense phase of the event was observed by SSM/I at 1531 UTC and by TMI at 1631 and 1808 UTC 5 December. This unusually short time sampling of a significant meteorological event at the midlatitudes makes this case a good challenge for satellite multisensor estimates, especially in view of future satellite missions aiming to provide more temporal sampling of PMW observations at global scales.

The NAW technique estimates have been applied to the Meteosat-7 infrared sequence, starting at 0000 UTC 4 December and ending at 0000 UTC 6 December. Six-hour-accumulated rainfall estimates show that rainfall peaks over the Antalya region both during the time interval of 1200–1800 UTC 5 December and 1800 UTC 5 December–0000 UTC 6 December (Fig. 6), consistent with the rain gauge reports and lightning activity. The NAW technique saturates to the highest rainfall value for a 6-h accumulation (i.e., 48 mm), while the Antalya gauge recorded 122 mm from 1200 to 1800 UTC 5 December.

Several methods have been proposed to merge passive microwave estimates (more quantitative) with geostationary IR data (higher spatial and temporal resolution), following different criteria (Todd et al. 2001; Turk et al. 2000). For the present case we applied the merging method proposed by Porcù et al. (2000), specifically designed for the analysis of intense rainfall events. This technique, based on the NAW processing of the Meteosat IR, uses PMW-estimated rainfall to provide the instantaneous rainfall rates in the areas labeled as “high precipitation” by the NAW processing. Both SSM/I and TMI rain estimates have been used for calibration during the most intense part of the event. While the TMI 2A12 standard rainfall product (Kummerow et al. 1996) was used for the two TRMM overpasses, a physical profile–based algorithm is used to retrieve rainfall from SSM/I data. This algorithm has been previously applied to the analysis of past Mediterranean storms, given its capability to also resolve deep convective cloud systems over land (Dietrich et al. 2000; Pinori et al. 2001).

The estimates for the three PMW overpasses are shown in Fig. 7. The heavy rainfall spots are evidenced by all three images, with the highest rain rates being 42, 28, and 35 mm h−1, respectively. The coastal location of the Antalya region poses problems for the microwave retrievals: usually coastal pixels are removed from rainfall maps because of the high difference in emissivity between land and sea, which can be misinterpreted as being an atmospheric signal by the PMW technique. In this case, for the TRMM overpasses, the rain area is clearly bounded by coastal lines (Figs. 7b and 7c), while IR observations indicate that the rain pattern extended over the sea. Despite this clear bias, we used the three PMW-derived fields to calibrate IR 6-h-accumulated rainfall. Comparison of the rainfall estimates after calibration (Fig. 8) with the estimates without calibration (Fig. 6) shows that the calibration resulted in an increase of the 6-h-accumulated rainfall during the time interval from 1200 to 1800 UTC 5 December, while it produced a decrease during the following 6-h interval. The calibration produced rainfall estimates closer to the Antalya rain gauge reports for the considered intervals (see Fig. 2), but the satellite technique is still not able to reproduce the highest rainfall peak that occurred in Antalya within 1200–1800 UTC (122 mm). On the other hand, the area of maximum precipitation as depicted in the satellite estimates coincides with the area where lightning has been recorded. Namely, comparison of Fig. 8 with Fig. 3 shows that during the period of 1200–1800 UTC 5 December the area of maximum estimated precipitation was narrower and apparently associated with intense and condensed in-space convective activity, while during the following 6 h the area of maximum precipitation was less intense but more spread as also depicted by the convective activity in the lightning location and in the flash-rate time plot in Fig. 4.

4. Model setup

MM5 (version 3.5) is a nonhydrostatic, primitive equation model using terrain-following coordinates (Dudhia 1993). Several physical parameterization schemes are available in the model for the boundary layer, radiative transfer, microphysics, and cumulus convection. MM5 has been run operationally at the National Observatory of Athens since 2000. For both the operational use and also for this case study the following schemes are selected: the Kain–Fritsch scheme (Kain and Fritsch 1993) for the convective parameterization, the scheme proposed by Schultz (1995) for the explicit microphysics, and the scheme by Hong and Pan (1996), which is also used by the National Centers for Environmental Prediction Global Forecasting System for the planetary boundary layer. The selection of the combination of the Kain–Fritsch scheme for convection and the Schultz scheme for explicit microphysics is based on the comparative study by Kotroni and Lagouvardos (2001). These authors had performed a comparison of various combinations of schemes for cases with important precipitation amounts over the eastern Mediterranean and showed that the combination of the Kain–Fritsch parameterization scheme with the highly efficient and simplified microphysical scheme proposed by Schultz (1995) provided the most skillful forecasts of accumulated precipitation for wintertime rain events in the area.

For this work two one-way-nested grids have been defined: grid 1 (with a grid spacing of 24 km), covering the major part of Europe, the Mediterranean, and the northern African borders, and grid 2 (8-km grid spacing), covering the western part of Turkey, the Aegean Sea, and a large part of the eastern Mediterranean. In the vertical dimension 30 unevenly spaced full sigma levels were selected.

Two simulations have been performed. For the first simulation, the ECMWF gridded analysis fields from 0000 UTC 4 December 2002 at a 6-h interval on a 0.5° latitude × 0.5° longitude horizontal grid increment have been used for initial and boundary conditions (CNTRL hereinafter). The second simulation has been motivated by the well-accepted argument that humidity is the least well analyzed field by global models (Ducrocq et al. 2002). For that reason a simple method of adjusting the humidity fields based on the use of the satellite precipitation estimates at 0000 UTC 4 December (calculated as explained in section 3b) has been applied (ASSIM hereinafter). For the present application a rain/no-rain flag has been assigned in each pixel of the domain used for the precipitation estimates. More specifically, given the ECMWF analysis at 0000 UTC 4 December, which is available at ∼0.5° × 0.5° resolution, and the satellite rainfall estimates, at the same time available at 0.25° × 0.25° resolution, the humidity adjustment is activated if at least half of the satellite grid boxes inside the ECMWF grid boxes (at least two boxes over four) have a “rain” flag. When this condition is fulfilled, the ECMWF relative humidity values inside the column (up to 300 hPa) over this specific ECMWF grid box are set to 100%. This method has been applied in a forecast mode for 16 cases over Greece and has proven its capability to improve the precipitation forecasts (Lagouvardos and Kotroni 2005). Both simulations lasted 72 h.

5. Simulation results

In this section the results of the simulations performed with MM5 are discussed. The t + 24 model output of mean sea level pressure from the model grid 1 (Fig. 9a) shows that the model was able to correctly reproduce the location of the low pressure system at 0000 UTC 5 December over Italy. It should be noted however that the model deepened the low by almost 4 hPa relative to the ECMWF analysis (Fig. 1c). Inspection of the 500-hPa geopotential height at the same time (Fig. 9b) shows a good agreement with the ECMWF analysis (Fig. 1d). Inspection of the model predictions at later times shows that although the model deepened the low more than was reported in the analysis, it has moved the system in good agreement with the analyses (and the observations), namely, in the area of interest (over the Antalya region).

To gain an insight on the reasons that produced the heavy amounts of precipitation in the Antalya region, a number of model-predicted fields from grid 2 are presented in the following (CNTRL run). Indeed, inspection of the equivalent potential temperature field at the 925-hPa level shows that the region of Antalya was influenced by the advection of warm and moist air masses since the beginning of the simulation. However, as time progresses the tongue of warm and moist air becomes narrower, and the discontinuity between the warm and moist air in the Antalya Gulf area with the less warm and drier air masses to the west becomes more important. This layer of warm and moisture-laden air masses is evident up to 850 hPa and the maximum of equivalent potential temperature values over the area are evident between 1200 and 1800 UTC 5 December (Figs. 10a and 10c), coinciding with the period of maximum precipitation. It should be also noted that during this period of the year the sea surface temperature in the area is relatively high, increasing thus moisture flux from the sea surface into the low-tropospheric air masses. The role of low-level moist air masses over the eastern Mediterranean region on the intensity of torrential rains has been also pointed out by Krichak et al. (2004). Inspection of the wind speed and direction reveals that an organized low-level jet (LLJ) was evident at the 925-hPa level over the maritime area between Cyprus and the southern coast of Turkey since 1200 UTC 4 December (not shown), which later on increased in magnitude and turned toward the Antalya Gulf at 1200 and 1800 UTC 5 December (Figs. 10b and 10d) when it reached its maximum speed (∼20 m s−1). The LLJ was evident up to the 850-hPa level (not shown). This LLJ advected the moist air into the region and the moist air stream progressed northwestward where a strong orographic enhancement of rainfall is expected as the flow impinged on the mountains west-northwest of Antalya.

The 6-h-accumulated precipitation fields at 1200 and 1800 UTC 5 December and 0000 UTC 6 December, provided by grid 2 are given in Fig. 11. Although, the surface network in the area of interest is not dense enough in order to further validate the precipitation forecasts, the comparison “by eye” of the spatial distribution of model-predicted rainfall at that time with the observations provided in Fig. 2 is quite favorable. The model predicts the moderate rainfall along the western Turkish coasts, in good agreement with the observations. As it concerns the area of interest, the maximum rainfall was predicted in the area of Antalya during the first two time intervals (up to 1200 and 1800 UTC), while during the last time interval (up to 0000 UTC 6 December) the area of maximum rainfall is displaced toward the northeast. The maximum 6-h-accumulated precipitation predicted by the CNTRL run was about 66 mm until 1200 UTC (Fig. 11a), overpredicting the observed 31-mm rainfall. During the next 6 h (until 1800 UTC, Fig. 11b) the predicted rain amount was 66 mm, underpredicting the observed 113-mm rainfall. During the last 6-h interval (up to 0000 UTC 6 December, Fig. 11c), the predicted amount is around 51 mm, an amount that is very close to the observed amount of 55 mm. The total predicted precipitation amount during the 18 h (from 0600 UTC 5 December to 0000 UTC 6 December) compares favorably with the observations (∼183 mm predicted and 208 mm observed), but the timing of the prediction is not successful because it distributes the rainfall almost equally in the three consecutive 6-h intervals.

The spatial distribution of model-predicted precipitation verifies favorably with the satellite estimates, especially after application of the PMW algorithm (Fig. 8), and with the lightning activity reported in Fig. 3. On the other hand, this comparison also shows that the model succeed in predicting the northwestward displacement of the area of precipitation maxima, but fails in timely producing the precipitation maxima up to 1800 UTC, as suggested by rain gauges, satellite rainfall estimates, and lightning activity.

The maximum of both the model-predicted and satellite-estimated precipitation is confined over the gulf and the plains of Antalya between the surrounding mountains, consistent with the extent of the reported lightning activity. Enhanced precipitation over the sea or the coast upwind of mountains seems to be strongly related to orography, as reported by Ogura et al. (1985) over Japan, Grossman and Durran (1984) over India, Caracena et al. (1979) over the United States, and Kotroni et al. (1999) over Greece. These authors have attributed the analyzed severe thunderstorms and deep convection midway between the mountain peaks and the shore to forced lifting of moist air directly to its lifting condensation level when this is near or below the top of the mountain barrier. Unfortunately, there is no rawinsonde station in the area in order to verify this idea. Nevertheless, this idea is supported by the model-predicted precipitation that produces its maxima not on the mountains surrounding Antalya but midway upstream the mountains.

At this point, the results of the sensitivity experiment in which the humidity adjustment was applied are discussed. As has been already mentioned in section 4 for this sensitivity experiment (ASSIM), the humidity of the ECMWF analysis at 0000 UTC 4 December has been adjusted based on the satellite rainfall estimates at the same time. Inspection of the model initial field of humidity (at 0000 UTC 4 December) in the low-tropospheric layers (namely, at 700 hPa) shows that in the CNTRL run the humidity over the western and southwestern Turkish coasts was on the order of 40%–60%, while in the ASSIM run it exceeded 80% in the same area (not shown). The argument that in the ASSIM run the humidity is better represented is supported by the fact that a number of surface stations in this area have reported rain at 0000 UTC 4 December. The 6-h-accumulated rainfall fields at 1200 and 1800 UTC 5 December and at 0000 UTC 6 December, provided by grid 2 of the ASSIM experiment, are given in Fig. 12. In general, except for the maximum over Antalya, the spatial distribution of rainfall between the two runs (CNTRL and ASSIM) is almost identical. At 1200 UTC the ASSIM experiment overestimates, as the CNTRL run also does, the precipitation over the Antalya region, producing almost the same maximum value (∼66 mm). Later on at 1800 UTC, the ASSIM run succeeds in producing a precipitation maximum much closer to that of the observations, while the spatial distribution of the precipitation field in comparison with the CNTRL run is very similar. Indeed, the rain gauge in Antalya reported 113 mm, the CNTRL run predicted a maximum ∼66 mm, while the ASSIM run produced a maximum of ∼91 mm. During the last reported 6-h period ending at 0000 UTC 6 December, the ASSIM run predicted a maximum of ∼61 mm, in good agreement with the observations and the CNTRL run. The above results, although based on one case, suggest that the application of the humidity adjustment technique is very promising. Moreover, the ASSIM experiment does not produce a general increase in the precipitation field over the whole domain, but instead it succeeds in increasing the precipitation maxima timing. These results support the idea that the improvement of the humidity field representation in the initial model fields is crucial for the improvement of the precipitation forecasts at later times. Further, the simplicity of the applied method makes it a successful tool for regional-scale studies of precipitation events. At this point it should be stressed that the above remarks have been only based on one case study. As shown by Lagouvardos and Kotroni (2005), the application of the humidity adjustment method on 16 cases over the eastern Mediterranean, produces statistically significant improvement of the quantitative precipitation forecasts.

6. Remarks—Prospects

In this paper, the storm that occurred over the Antalya region on 5 December 2002 has been studied. The analysis was based on the model results provided by MM5 but also the observations provided by satellites and lightning detection sensors. The satellite data include those provided by TMI, SSM/I, and Meteosat.

The sampling of the event by satellites was exceptionally good (three overpasses during maximum intensity), making the case very suitable for satellite multisensor rainfall estimation techniques. These estimates were able to delineate areas where peaks of 6-h-accumulated rainfall occurred, but given the lack of a reliable and significant ground validation dataset, no evaluation on the capability to quantitatively estimate rainfall peaks was given. The application of the PMW–IR technique for calibration of the precipitation estimates resulted in a more realistic retrieval of precipitation over the area, especially as it concerns the timing of the precipitation maxima.

The lightning activity associated with this storm has been analyzed based on the Met Office ATD system as well as the TRMM LIS data. The time evolution of the lightning activity over the area showed that during the period at that maximum precipitation was observed the electrification phenomena associated with the storm were intense and the activity almost stationary over the Antalya area, while later on the activity moved east and weakened significantly. The two overpasses of TRMM during the maximum of the storm intensity also permitted estimation of the ratio of the cloud to ground over the total amount of lightning that ranged from 0.21 to 0.33.

The analysis of the model results showed that the model was able to correctly reproduce the location of the precipitation maxima in the area of Antalya. The subjective comparison with in situ observations and the satellite rainfall estimates was very promising. Indeed, the precipitation maxima occurred when the low-level flow of warm and moist air has been intensified and impinged almost perpendicularly on the mountains west of Antalya. Although the model simulated well the total precipitation amount during the event, it was not successful in distributing the precipitation correctly in time (at 6-h intervals). For that reason a sensitivity test has been performed, where the humidity field in the model initial conditions has been adjusted based on the satellite-estimated rainfall. This experiment showed that with the humidity adjustment the model reproduced more correctly the maxima of precipitation in the time and area of interest, without increasing in general the precipitation field in the whole model domain.

This paper presented a synthesis of observational and model data that can be used synergistically for the analysis of rainfall events. Further experimentation is needed in order to prove also the effectiveness of the humidity adjustment technique applied on model initial conditions for the improvement of the quantitative precipitation estimates.

Acknowledgments

This work has been jointly financed by the European Union (75%) and the Greek Ministry of Development (25%) in the framework of the program “Competitiveness—Promotion of Excellence in Technological Development and Research—Excellence in Research Centers, Action 3.3.1,” (MIS64563) and by the Greek General Secretariat for Research and Technology (NATO/CCMS Pilot Study). The Italian team has been funded by the Italian National Group for Prevention from Hydro-Geological Disasters (GNDCI), and within the framework of EURAINSAT, a shared-cost project (contract EVG1-2000-00030) cofunded by the Research DG of the European Commission within the RTD activities (5th Framework Program).

REFERENCES

  • Alpert, P., , B. U. Neeman, , and Y. Shay-El, 1990: Climatological analysis of Mediterranean cyclones using ECMWF data. Tellus, 42 , 6577.

    • Search Google Scholar
    • Export Citation
  • Alpert, P., , M. Tzidulko, , and D. Izigsohn, 1999: A shallow short-lived meso-beta cyclone over the gulf of Antalya, eastern Mediterranean. Tellus, 51A , 249262.

    • Search Google Scholar
    • Export Citation
  • Buzzi, A., , N. Tartaglione, , and P. Malguzzi, 1998: Numerical simulations of the 1994 Piedmont flood: Role of orography and moist processes. Mon. Wea. Rev, 126 , 23692383.

    • Search Google Scholar
    • Export Citation
  • Caracena, F., , R. A. Maddox, , L. R. Hoxit, , and C. F. Chappell, 1979: Mesoanalysis of the Big Thompson storm. Mon. Wea. Rev, 107 , 117.

  • Christian, H. J., and Coauthors, 1999: The lightning imaging sensor. Proc. 11th Int. Conf. on Atmospheric Electricity, Huntsville, AL, ICAE, 746–749.

  • Cummins, K. L., , M. J. Murphy, , E. A. Bardo, , W. L. Hiscox, , R. B. Pyle, , and A. E. Pifer, 1998: A combined TOA/MDF technology upgrade of the U.S. National Lightning Detection Network. J. Geophys. Res, 103 , D8. 90359044.

    • Search Google Scholar
    • Export Citation
  • Dietrich, S., , C. Adamo, , A. Mugnai, , and R. Bechini, 2000: A physical profile-based multisensor precipitation retrieval. Proc. 2000 EUMETSAT Meteorological Satellite Data Users' Conf., Bologna, Italy, EUMETSAT, Vol. EUM P29, 443–449.

  • Ducrocq, V., , D. Ricard, , J-P. Lafore, , and F. Orain, 2002: Storm-scale numerical rainfall prediction for five precipitating events over France: On the importance of the initial humidity field. Wea. Forecasting, 17 , 12361256.

    • Search Google Scholar
    • Export Citation
  • Dudhia, J., 1993: A non-hydrostatic version of the Penn State/NCAR Mesoscale Model: Validation tests and simulation of an Atlantic cyclone and cold front. Mon. Wea. Rev, 121 , 14931513.

    • Search Google Scholar
    • Export Citation
  • Ferretti, R., , S. Low-Nam, , and R. Rotunno, 2000: Numerical simulations of the Piedmont flood of 4-6 November 1994. Tellus, 52 , 162180.

    • Search Google Scholar
    • Export Citation
  • Grossman, R. L., , and D. R. Durran, 1984: Interaction of low-level flow with the western Ghat Mountains and offshore convection in the summer monsoon. Mon. Wea. Rev, 112 , 652672.

    • Search Google Scholar
    • Export Citation
  • Holt, M. A., , P. J. Hardaker, , and G. P. McLelland, 2001: A lightning climatology for Europe and the UK, 1990-99. Weather, 56 , 290296.

  • Hong, S-Y., , and H-L. Pan, 1996: Nonlocal boundary layer vertical diffusion in a medium-range forecast model. Mon. Wea. Rev, 124 , 23222339.

    • Search Google Scholar
    • Export Citation
  • Jansà, A., , A. Genovés, , and J. A. García-Moya, 2000: Western Mediterranean cyclones and heavy rain. Part 1: Numerical experiment concerning the Piedmont flood case. Meteor. Appl, 7 , 323333.

    • Search Google Scholar
    • Export Citation
  • Kain, J. S., , and J. M. Fritsch, 1993: Convective parameterization for mesoscale models: The Kain-Fritsch scheme. The Representation of Cumulus in Numerical Models, Meteor. Monogr., No. 46, Amer. Meteor. Soc., 165–177.

  • Kotroni, V., , and K. Lagouvardos, 2001: Precipitation forecast skill of different convective parameterization and microphysical schemes: Application for the cold season over Greece. Geophys. Res. Lett, 28 , 19771980.

    • Search Google Scholar
    • Export Citation
  • Kotroni, V., , K. Lagouvardos, , G. Kallos, , and D. Ziakopoulos, 1999: Severe flooding over central and southern Greece associated with pre-cold frontal orographic lifting. Quart. J. Roy. Meteor. Soc, 125 , 967991.

    • Search Google Scholar
    • Export Citation
  • Krichak, S. O., , P. Alpert, , and M. Dayan, 2004: Role of atmospheric processes associated with Hurricane Olga in the December 2001 floods in Israel. J. Hydrometeor, 5 , 12591270.

    • Search Google Scholar
    • Export Citation
  • Kummerow, C. W., , W. S. Olson, , and L. Giglio, 1996: A simplified scheme for obtaining precipitation and vertical hydrometeors profiles from passive microwave sensors. IEEE Trans. Geosci. Remote Sens, 34 , 12131232.

    • Search Google Scholar
    • Export Citation
  • Kummerow, C., , W. Barnes, , T. Kozu, , J. Shiue, , and J. Simpson, 1998: The Tropical Rainfall Measuring Mission (TRMM) sensor package. J. Atmos. Oceanic Technol, 15 , 809817.

    • Search Google Scholar
    • Export Citation
  • Kurz, M., , and A. D. Fontana, 2004: A case of cyclogenesis over the western Mediterranean Sea with extraordinary convective activity. Meteor. Appl, 11 , 97113.

    • Search Google Scholar
    • Export Citation
  • Lagouvardos, K., , and V. Kotroni, 1999: Aspects of the 25-26 March 1998 storm over Greece. Weather, 54 , 154160.

  • Lagouvardos, K., , and V. Kotroni, 2000: Use of METEOSAT water-vapour images for the diagnosis of a vigorous stratospheric intrusion over Central Mediterranean. Meteor. Appl, 7 , 205210.

    • Search Google Scholar
    • Export Citation
  • Lagouvardos, K., , and V. Kotroni, 2005: Improvement of high resolution weather forecasts through humidity adjustment, based on satellite data. Quart. J. Roy. Meteor. Soc, 131 , 26952712.

    • Search Google Scholar
    • Export Citation
  • Lagouvardos, K., , V. Kotroni, , S. Dobricic, , S. Nickovic, , and G. Kallos, 1996: On the storm of 21-22 October 1994 over Greece: Observations and model results. J. Geophys. Res, 101 , D21. 2621726226.

    • Search Google Scholar
    • Export Citation
  • Lagouvardos, K., , V. Kotroni, , S. Nickovic, , D. Jovic, , and G. Kallos, 1999: Observations and model simulations of a winter sub-synoptic vortex over the central Mediterranean. Meteor. Appl, 6 , 371383.

    • Search Google Scholar
    • Export Citation
  • Lee, A. C. L., 1986: An experimental study of the remote location of lightning flashes using VLF arrival time difference technique. Quart. J. Roy. Meteor. Soc, 112 , 203229.

    • Search Google Scholar
    • Export Citation
  • Maheras, P., , H. A. Flocas, , C. Anagnostopoulou, , and I. Patrikas, 2002: On the vertical structure of composite surface cyclones in the Mediterranean region. Theor. Appl. Climatol, 71 , 3–4. 199217.

    • Search Google Scholar
    • Export Citation
  • Negri, A. J., , R. F. Adler, , and P. J. Wetzel, 1984: Rain estimation from satellite: An examination of the Griffith–Woodley technique. J. Climate Appl. Meteor, 23 , 102116.

    • Search Google Scholar
    • Export Citation
  • Ogura, Y., , T. Asai, , and K. Dohi, 1985: A case study of heavy precipitation event along the Baiu front in northern Kyushu, 23 July 1982: Nagasaki heavy rainfall. J. Meteor. Soc. Japan, 63 , 883900.

    • Search Google Scholar
    • Export Citation
  • Pinori, S., and Coauthors, 2001: MW/IR satellite data for a combined rainfall estimation of Piemonte and Valle d'Aosta 2000 flood event. Proc. Third EGS Plinius Conf. on Mediterranean Storms, Baja Sardinia, Italy, European Geophysical Society, 161–170.

  • Porcù, F., , F. Prodi, , S. Dietrich, , A. Mugnai, , and R. Bechini, 2000: Multisensor estimation of severe rainfall events. Proc. 2000 Eumetsat Meteorological Satellite Data Users' Conf., Darmstadt, Germany, EUMETSAT, Vol. EUM P29, 371–378.

  • Porcù, F., , C. Caracciolo, , and F. Prodi, 2003: Cloud systems leading to flood events in Europe: Overview and classification. Meteor. Appl, 10 , 217227.

    • Search Google Scholar
    • Export Citation
  • Schultz, P., 1995: An explicit cloud physics parameterization for operational numerical weather prediction. Mon. Wea. Rev, 123 , 33313343.

    • Search Google Scholar
    • Export Citation
  • Todd, M. C., , C. Kidd, , D. Kniverton, , and T. J. Bellerby, 2001: A combined satellite infrared and passive microwave technique for estimation of small-scale rainfall. J. Atmos. Oceanic Technol, 18 , 742755.

    • Search Google Scholar
    • Export Citation
  • Trigo, I. F., , G. R. Bigg, , and T. D. Davies, 2002: Climatology of cyclogenesis mechanisms in the Mediterranean. Mon. Wea. Rev, 130 , 549569.

    • Search Google Scholar
    • Export Citation
  • Tripoli, G. J., , G. Panegrossi, , S. Dietrich, , A. Mugnai, , E. A. Smith, , and F. Siccardi, 2001: A numerical study of the 13-16 October, 2000 Piemonte flood: Comparison of meteorological conditions to two other contrasting fall season flooding events in northern Italy. Proc. Third EGS Plinius Conf. on Mediterranean Storms, Baja Sardinia, Italy, European Geophysical Society, 115–119.

  • Tsidulko, M., , and P. Alpert, 2001: Synergism of upper-level potential vorticity and mountains in Genoa lee cyclogenesis: A numerical study. Meteor. Atmos. Phys, 78 , 261285.

    • Search Google Scholar
    • Export Citation
  • Turk, F. J., , S. Chi-Sann Liou, , R. Qiu, , M. Scofield, , B. Ba, , and A. Gruber, 2000: Capabilities and characteristics of rainfall estimates from geostationary- and geostationary + microwave-based satellite techniques. Preprints. Symp. on Precipitation Extremes: Prediction, Impacts, and Responses, Albuquerque, NM, Amer. Meteor. Soc., 191–194.

    • Search Google Scholar
    • Export Citation
Fig. 1.
Fig. 1.

(a) Analysis of mean sea level pressure at 2-hPa intervals from the ECMWF at 0000 UTC 4 Dec. Equivalent potential temperature values (at 850 hPa) greater than 308 K are shaded at 2-K interval. (b) Analysis of the 500-hPa-level geopotential height at 40-m interval. Shading denotes wind speed at 300-hPa level contoured at 10 m s−1 interval (only values exceeding 40 m s−1 are plotted). (c) and (e) As in (a), but at 0000 and 1800 UTC 5 Dec, respectively; (d) and (f) as in (b), but for 0000 and 1800 UTC 5 Dec, respectively.

Citation: Journal of Applied Meteorology and Climatology 45, 4; 10.1175/JAM2347.1

Fig. 2.
Fig. 2.

The 6-h-accumulated rainfall (mm) from the surface network reports, ending at 1200 and 1800 UTC 5 Dec 2002 and 0000 UTC 6 Dec 2002. Respective values are separated by a slash. Antalya is denoted by an asterisk.

Citation: Journal of Applied Meteorology and Climatology 45, 4; 10.1175/JAM2347.1

Fig. 3.
Fig. 3.

Geolocations of the CG flashes in 3-h intervals as recorded by the ATD system for the period from 0600 UTC 5 Dec 2002 to 0600 UTC 6 Dec 2002.

Citation: Journal of Applied Meteorology and Climatology 45, 4; 10.1175/JAM2347.1

Fig. 4.
Fig. 4.

(a) Time series of the CG flash rate and (b) maximum of flash density, calculated in 5-min intervals for the period from 0600 UTC 5 Dec 2002 to 0600 UTC 6 Dec 2002. Flash density was computed by 0.2° lat × 0.2° lon.

Citation: Journal of Applied Meteorology and Climatology 45, 4; 10.1175/JAM2347.1

Fig. 5.
Fig. 5.

TMI 85-GHz horizontal polarization images over Antalya region during (a) the first overpass at 1631 UTC 5 Dec and (b) the second overpass at 1808 UTC the same day. Examples of two flashes recorded by LIS at (c) 1807:55 and (d) 1808:27 UTC (see text for details) are shown. (e) The shades of gray coded time series of the maximum of optical magnitude per LIS frame recorded during the flash shown in (c); (f) as in (e), but for the flash shown in (d). Triangle in (f) shows the time of the ATD fix recorded during the same flash. Its location is indicated at the intersection of the horizontal and vertical solid lines in (d).

Citation: Journal of Applied Meteorology and Climatology 45, 4; 10.1175/JAM2347.1

Fig. 6.
Fig. 6.

The 6-h-accumulated rainfall ending at (a) 1800 UTC 5 Dec 2002 and at (b) 0000 UTC 6 Dec 2002 as calculated by the NAW method.

Citation: Journal of Applied Meteorology and Climatology 45, 4; 10.1175/JAM2347.1

Fig. 7.
Fig. 7.

PMW rain-rate estimates (mm h−1) from the (a) SSM/I overpass (1531 UTC), (b) TRMM overpass (1631 UTC), and (c) TRMM overpass (1808 UTC).

Citation: Journal of Applied Meteorology and Climatology 45, 4; 10.1175/JAM2347.1

Fig. 8.
Fig. 8.

The 6-h-accumulated rainfall ending at (a) 1800 UTC 5 Dec 2002 and at (b) 0000 UTC 6 Dec 2002 as calculated by the PMW–IR method.

Citation: Journal of Applied Meteorology and Climatology 45, 4; 10.1175/JAM2347.1

Fig. 9.
Fig. 9.

(a) Mean sea level pressure at 2-hPa intervals and (b) 500-hPa geopotential height at 40-m interval. Shading denotes wind speed at 300-hPa level contoured at 10 m s−1 interval (only values exceeding 40 m s−1 are plotted) valid at 0000 UTC 5 Dec 2002, as predicted by MM5 grid 1.

Citation: Journal of Applied Meteorology and Climatology 45, 4; 10.1175/JAM2347.1

Fig. 10.
Fig. 10.

(a) The 925-hPa equivalent potential temperature (at 2-K intervals; values greater than 308 K are shaded), (b) The 925-hPa wind speed (contoured at 2 m s−1 interval; only values greater than 10 m s−1 are shown) from MM5 grid 2 (CNTRL run), valid at 1200 UTC 5 Dec 2002. (c) As in (a), but at 1800 UTC 5 Dec, (d) as in (b), but at 1800 UTC 5 Dec, (e) as in (a) but at 0000 UTC 6 Dec, and (f) as in (b), but at 0000 UTC 6 Dec.

Citation: Journal of Applied Meteorology and Climatology 45, 4; 10.1175/JAM2347.1

Fig. 11.
Fig. 11.

The 6-h-accumulated precipitation (at 10-mm interval) valid at (a) 1200 UTC 5 Dec 2002, (b) 1800 UTC 5 Dec 2002, and (c) 0000 UTC 6 Dec 2002 as predicted by MM5 grid 2 (CNTRL run). The mountain barrier around Antalya is denoted by a thick line.

Citation: Journal of Applied Meteorology and Climatology 45, 4; 10.1175/JAM2347.1

Fig. 12.
Fig. 12.

The 6-h-accumulated precipitation (at 10-mm interval) valid at (a) 1200 UTC 5 Dec 2002, (b) 1800 UTC 5 Dec 2002, and (c) 0000 UTC 6 Dec 2002 as predicted by MM5 grid 2 (ASSIM run).

Citation: Journal of Applied Meteorology and Climatology 45, 4; 10.1175/JAM2347.1

Save