• Abdalla, S., , Bidlot J-R. , , and Janssen P. , 2005: Assimilation of ERS and Envisat wave data at ECMWF. Proc. 2004 Envisat and ERS Symp., Publication SP-572, Salzburg, Austria, ESA, ESTEC, CD-ROM.

  • Alves, J-H. G. M., , and Young I. R. , 2003: On estimating extreme wave heights using combined Geosat, TOPEX/Poseidon and ERS-1 altimeter data. Appl. Ocean Res., 25 , 167186.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Aouf, L., , Lefèvre J-M. , , and Hauser D. , 2006: Assimilation of directional wave spectra in the wave model WAM: An impact study from synthetic observations in preparation for the SWIMSAT satellite mission. J. Atmos. Oceanic Technol., 23 , 448463.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • AVISO, 1996: AVISO user handbook: Merged TOPEX/POSEIDON products. 3.0 ed. CLS-CNES AVI-NT-02-101, 201 pp.

  • Bentamy, A., , Ayina H-L. , , Queffeulou P. , , Croize-Fillon D. , , and Kerbaol V. , 2007: Improved near real time surface wind resolution over the Mediterranean Sea. Ocean Sci., 3 , 259271.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Brody, L. R., , and Nestor M. J. R. , 1980: Handbook for forecasters in the Mediterranean, Part 2: Regional forecasting aids for the Mediterranean Basin. Naval Environmental Prediction Research Facility Tech. Rep. TR 80-10, 178 pp.

  • Caires, S., , and Sterl A. , 2005a: A new nonparametric method to correct model data: Application to significant wave height from the ERA-40 Re-Analysis. J. Atmos. Oceanic Technol., 22 , 443459.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Caires, S., , and Sterl A. , 2005b: 100-year return value estimates for ocean wind speed and significant wave height from the ERA-40 data. J. Climate, 18 , 10321048.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • CERSAT, 1996: Altimeter and microwave radiometer ERS products user manual, version 2.2. CERSAT, IFREMER C2-MUT-A-01-IF, 136 pp.

  • Chen, G., , Chapron B. , , Ezraty R. , , and Vandemark D. , 2002: A global view of swell and wind sea climate in the ocean by satellite altimeter and scatterometer. J. Atmos. Oceanic Technol., 19 , 18491859.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • ESA, 2002: Envisat RA-2/MWR product handbook. Product Document PO-TN-ESR-RA-0050, version 1.2, 185 pp.

  • Greenslade, D. J. M., , and Young I. R. , 2004: Background errors in a global wave model determined from altimeter data. J. Geophys. Res., 109 .C09007, doi:10.1029/2004JC002324.

    • Search Google Scholar
    • Export Citation
  • Greenslade, D. J. M., , and Young I. R. , 2005: The impact of altimeter sampling patterns on estimates of background errors in a global wave model. J. Atmos. Oceanic Technol., 22 , 18951917.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Medatlas Group, 2004: Wind and Wave Atlas of the Mediterranean Sea. WEAO Research Cell, 420 pp.

  • Mitchum, G. T., , Hancock D. W. III, , Hayne G. S. , , and Vandemark D. C. , 2004: Blooms of σ0 in the TOPEX radar altimeter data. J. Atmos. Oceanic Technol., 21 , 12321244.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Naval Oceanographic Office/NOAA Laboratory for Satellite Altimetry, 2002: GEOSAT follow-on GDR user’s handbook. NOAA/NESDIS/ORA:E/RA31, 20 pp.

  • Picot, N., , Case K. , , Desai S. , , and Vincent P. , 2003: AVISO and PODAAC user handbook: IGDR and GDR Jason products. AVISO SMM-MU-M5-OP-13184-CN, PODAAC JPL D-21352, 107 pp.

  • Queffeulou, P., 2004: Long-term validation of wave height measurements from altimeters. Mar. Geod., 27 , 495510.

  • Quilfen, Y., , Chapron B. , , Collard F. , , and Serre M. , 2004: Calibration validation of an altimeter wave period model and application to TOPEX Poseidon and Jason-1 altimeters. Mar. Geod., 27 , 535549.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sasaki, W., , Iwasaki S. I. , , Matsuura T. , , Iizuka S. , , and Watabe I. , 2005: Changes in wave climate off Hiratsuka, Japan, as affected by storm activity over the western North Pacific. J. Geophys. Res., 110 .C09008, doi:10.1029/2004JC002730.

    • Search Google Scholar
    • Export Citation
  • Skandrani, C., , Lefevre J-M. , , and Queffeulou P. , 2004: Impact of multisatellite altimeter data assimilation on wave analysis and forecast. Mar. Geod., 27 , 511533.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sterl, A., , and Caires S. , 2005: Climatology, variability and extrema of ocean waves: The web-based KNMI/ERA-40 wave atlas. Int. J. Climatol., 25 , 963977.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sotillo, M. G., , Ratsimandresy A. W. , , Carretero J. C. , , Bentamy A. , , Valero F. , , and Gonzales-Rouco F. , 2005: A high-resolution 44-year atmospheric hindcast for the Mediterranean Basin: Contribution to the regional improvement of global reanalysis. Climate Dyn., 25 , 219236.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wolf, J., , and Woolf D. K. , 2006: Waves and climate change in the north-east Atlantic. Geophys. Res. Lett., 33 .L06604, doi:10.1029/2005GL025113.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Map of the Mediterranean Sea subbasins.

  • View in gallery

    ERS-1 ground tracks during the 3-day repeat cycle of phase A. The arrow indicates the particular track of interest.

  • View in gallery

    (a) Individual ERS-1 SWH measurements every 3 days along the particular track of Fig. 1 from 15 Sep to 15 Dec 1991. (b) Three-month averages of ERS-1 SWH measurements along the particular track as a function of latitude for the initial 3-day (solid line) and 9-day samples (circles, squares, or crosses).

  • View in gallery

    Thirteen-year along-track mean SWH for January.

  • View in gallery

    (a)–(d) Maps of 13-yr along-track seasonal mean SWH and (e) winter map zoom.

  • View in gallery

    Maps of 13-yr along-track seasonal SWH standard deviation.

  • View in gallery

    Mediterranean basin time series of (a) number of data for each altimeter; (b) total of the missions; (c), (e) mean value; and (d) std dev of SWH measurements of the 6 altimeters over 14 yr on a monthly basis.

  • View in gallery

    Mediterranean subbasins: SWH (a) distributions and (b) cumulative distributions from the 6 altimeters over 14 yr.

  • View in gallery

    Comparisons of SWH annual cycles from buoy (crosses) and altimeter (circles) measurements. (a) Monthly mean value, (b) normalized mean value, and (c) scatter index.

  • View in gallery

    Monthly annual cycle of SWH, over 14 yr, for (a), (c), (e) western and (b), (d), (f) eastern Mediterranean subbasins. (a), (b) Monthly absolute and (c), (d) normalized mean values. (e), (f) Monthly scatter index.

  • View in gallery

    SWH seasonal (solid lines) time series and mean annual cycle (dashed lines) for the western Mediterranean subbasins.

  • View in gallery

    SWH seasonal (solid lines) time series and mean annual cycle (dashed lines) for the eastern Mediterranean subbasins.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 73 73 22
PDF Downloads 77 77 26

Analysis of Wave Height Variability Using Altimeter Measurements: Application to the Mediterranean Sea

View More View Less
  • 1 Laboratoire d’Océanographie Spatiale, Institut Français de Recherche pour l’Exploitation de la Mer, Plouzané, France
© Get Permissions
Full access

Abstract

Altimeter significant wave height (SWH) measurement data from six satellite missions covering 14 yr were analyzed over the Mediterranean Sea. First, data correction and screening were performed using the same method for the six altimeters [European Remote Sensing Satellites (ERS-1 and ERS-2), Ocean Topography Experiment (TOPEX), Geosat Follow-On, Jason, and Environmental Satellite (Envisat)]. The data from the TOPEX and Jason missions enabled the construction of seasonal maps of along-track SWH mean values and standard deviations. These reveal the regional short-scale sea state features associated with the specific meteorological patterns of the various geographical basins. Time series of monthly SWH mean values and standard deviations from each satellite and over the whole Mediterranean Sea were calculated and seen to be in good agreement, thus demonstrating interannual variability. The six altimeter missions used together enable the investigation of the monthly annual cycle at the short scales of the various subbasins. Significant differences are observed between the western and eastern parts of the Mediterranean Sea. The annual SWH cycle changes in both shape and amplitude depending on the subbasin. Analysis of the seasonal interannual variability confirms the existence of some degree of independence between the subbasins. Thanks to multisatellite missions and homogeneous corrections of the altimeter data, SWH time and space characteristics were able to be obtained at regional short scales. These results are independent of numerical wind and wave models. This method can be applied to any geographical region.

Corresponding author address: Pierre Queffeulou, Laboratoire d’Océanographie Spatiale, BP 70, 29280 Plouzané, France. Email: pierre.queffeulou@ifremer.fr

Abstract

Altimeter significant wave height (SWH) measurement data from six satellite missions covering 14 yr were analyzed over the Mediterranean Sea. First, data correction and screening were performed using the same method for the six altimeters [European Remote Sensing Satellites (ERS-1 and ERS-2), Ocean Topography Experiment (TOPEX), Geosat Follow-On, Jason, and Environmental Satellite (Envisat)]. The data from the TOPEX and Jason missions enabled the construction of seasonal maps of along-track SWH mean values and standard deviations. These reveal the regional short-scale sea state features associated with the specific meteorological patterns of the various geographical basins. Time series of monthly SWH mean values and standard deviations from each satellite and over the whole Mediterranean Sea were calculated and seen to be in good agreement, thus demonstrating interannual variability. The six altimeter missions used together enable the investigation of the monthly annual cycle at the short scales of the various subbasins. Significant differences are observed between the western and eastern parts of the Mediterranean Sea. The annual SWH cycle changes in both shape and amplitude depending on the subbasin. Analysis of the seasonal interannual variability confirms the existence of some degree of independence between the subbasins. Thanks to multisatellite missions and homogeneous corrections of the altimeter data, SWH time and space characteristics were able to be obtained at regional short scales. These results are independent of numerical wind and wave models. This method can be applied to any geographical region.

Corresponding author address: Pierre Queffeulou, Laboratoire d’Océanographie Spatiale, BP 70, 29280 Plouzané, France. Email: pierre.queffeulou@ifremer.fr

1. Introduction

The wind and wave conditions over the Mediterranean Sea are characterized by particularly high space and time variability. There are several reasons for this. First, the Mediterranean Sea is located at the boundary between three typical meteorological weather patterns: the oceanic regime of the northeast Atlantic Ocean, dominated by both the position and movement of the Azores high pressure area and the low pressure systems moving across the northeast Atlantic Ocean; the warm continental regime resulting from the deserts of North Africa; and the continental regime associated with the Eurasian landmass and southern Asia. These three different meteorological regimes demonstrate varying levels of influence depending on the geographical location and on the time of year. Second, the Mediterranean Sea is separated into several distinct geographical basins. Figure 1 shows these nine subbasins, selected according to the sea forecast areas as established by the meteorological offices (Brody and Nestor 1980). The western Mediterranean, Thyrrenian, Adriatic, and Aegean Seas are nearly enclosed seas, separated by narrow straits. A third fundamental characteristic is that these basins are surrounded by large chains of mountains with many coastal valleys opening onto the sea. These features perform as barriers to airflows while inducing both funneling effects and katabatic winds (cold air flowing downward from the mountain top). Thus it is seen that meteorological and geographical features induce large variability in the winds with strong seasonal changes and variations at both short and basin scales. These specific Mediterranean wind regimes, together with the presence of numerous islands, induce an even larger complexity in describing and monitoring of the resulting sea states.

To assist in determining regional wave statistics, data from wave buoys, numerical wave models, and satellites are readily available. The number of wave buoys over the oceans continues to increase. These now include operational buoys from sources such as the U.S. National Data Buoy Center, the Canadian Marine Environmental Data Services, the European buoy network maintained by Ireland, the Met Office (U.K.), Météo France, Spain, Italy, and Greece. These buoys, however, are often located in coastal regions and are not uniformly distributed in space. In addition, because of the difficulties in maintaining buoys, long time series (several years) are rare. For this reason buoy data are only useful for wave climates in certain locations (see, e.g., Sasaki et al. 2005) and are not fully adapted to estimates of wave descriptions over subbasins of the Mediterranean Sea.

During the last few years the quality of numerical wave model analyses has improved considerably thanks to improved wind estimates and to the development of assimilation techniques, using buoy and satellite [altimeter, scatterometer, and synthetic aperture radar (SAR) images] measurements (Abdalla et al. 2005; Skandrani et al. 2004; Aouf et al. 2006). The skill of altimeter assimilation in wave models, however, is under discussion (Greenslade and Young 2004, 2005). An important question is the quality of numerical wave models. For instance, the 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40) produces both underestimation of high significant wave height (SWH) and problems of time heterogeneity that must be corrected (Caires and Sterl 2005a) before application to wave climate studies (Caires and Sterl 2005b; Sterl and Caires 2005; Wolf and Woolf 2006).

The problems observed in wave models at global scale are exacerbated by conditions particular to enclosed seas, such as the Mediterranean Sea. These include the high time variability and short spatial scales of the winds, the effect of short fetches, and the presence of many islands. In recent years a particular effort has been made to assess the climate variability of the Mediterranean Sea. To this end, European Union (EU)-funded projects were established to generate long-term meteorological and oceanographical time series. High-resolution wind fields were produced from atmospheric models (Sotillo et al. 2005) within the Hindcast of Dynamic Processes of the Ocean and Coastal Areas of Europe (HIPOCAS) project. Blended wind products using both atmospheric model and satellite data are currently in development (Bentamy et al. 2007) through the Mediterranean Forecasting System Toward Environmental Predictions (MFSTEP) project. An atlas of waves for the Mediterranean Sea is now available, using a 10-yr (1992–2002) archive of ECMWF wind and wave model data (Medatlas Group 2004). This work required an intensive effort of model (wind and wave) output calibration and correction, largely using altimeter data from the European Remote Sensing Satellites (ERS) and Ocean Topography Experiment (TOPEX). In general, significant underestimation of model-derived wind speed and SWH were observed.

The third source of data used in deriving regional wave characteristics is the direct use of satellite measurements, mainly from SAR and altimetry. SAR images present the benefit of providing information on both wave height and spectrum, but the time and space coverage is still too sparse over the Mediterranean Sea. In addition, the relatively low resolution of the measurements fails to properly capture short wave systems.

For regional SWH statistics, the altimeter is very useful because almost 15 yr of continuous data are presently available, commencing with the launch of ERS-1 in 1991. The main motivation for this paper is that buoys rarely produce long continuous time series and have a very sparse geographical distribution, whereas altimeters provide global, continuous, and long-term measurements. A strong interest stems also from the along-track altimeter high-resolution sampling (on the order of 6 km), which enables local measurements of SWH gradients induced by fetch or sheltering effects. Nevertheless, the altimeter measurements have several disadvantages. First, the altimeter footprint is very narrow, providing a measurement along the ground track only at the nadir location. Second, the measurement is limited to significant wave height and wind speed (not direction) with no information on the spectrum, although the possibility of a mean wave period estimate has been investigated (Quilfen et al. 2004). It was also demonstrated that the use of simultaneous wave height and wind speed information enables the analysis of the space distribution of wind sea and swell over the global oceans (Chen et al. 2002). Third, the altimeter time and space sampling scheme is a compromise between revisiting the same ground track frequently, as for the 3-day repeat cycle orbit of ERS-1 during the commissioning phase with a coarse spatial resolution (about 8° in longitude intertrack separation at 40°N), and revisiting at a longer time period (35 day for ERS-2) with a better spatial sampling. The various altimetric missions exhibit a high variability in the sample: ERS-1 (3, 168, and 35 days), TOPEX/Poseidon (10 days), Geosat Follow-On (GFO; 17 days), ERS-2 (35 days), Jason-1 (10 days), and Environmental Satellite (Envisat; 35 days).

In this paper both along-track high-resolution data and multisatellite altimeter sampling are used in precisely determining the wave height description of the Mediterranean Sea. The altimeter data and processing are described in the first part. The second part takes advantage of the high-resolution along-track sampling to investigate the seasonal spatial variability of SWH along the TOPEX and Jason ground tracks (almost 14 yr). Then, in the third part, TOPEX and Jason data are complemented with altimeter data from ERS-1, ERS-2, Geosat Follow-On, and Envisat. These data are then used to estimate the seasonal annual cycle and the interannual variability of SWH, both on the scale of the full Mediterranean basin, including the Black Sea, on a monthly basis and for the various aforementioned subbasins.

2. Data and processing

The altimeter data are retrieved from the Geophysical Data Records (GDRs), or the equivalent, distributed by the various space agencies. Details of the time characteristics of the various altimeter datasets are given in Table 1. The ERS-1 data are the European Space Agency (ESA) Radar altimeter Ocean Product (OPR) described in CERSAT (1996) for phases C and G. The ERS-1 phases A, B, E, and F are not available in the OPR format; the reprocessing is still ongoing. For consistency in the data, only phases C and G are used in the statistical analysis. Nevertheless, Ocean Intermediate Products (OIPs) from ERS-1 phase A are able to be used in section 3 to illustrate the altimeter sampling effect. ERS-2 data are the OPR (CERSAT 1996) covering the entire mission commencing April 1995. After 22 June 2003, the ERS-2 coverage was seriously reduced due to the failure of the onboard tape recorder. TOPEX/Poseidon data are from the Merged Geophysical Data Record (M-GDR) described in AVISO (1996), covering the whole mission from September 1992 to October 2005. Jason data are the Jason GDRs (Picot et al. 2003) from January 2002 to December 2005. Geosat Follow-On data are the GDR (Naval Oceanographic Office/NOAA Laboratory for Satellite Altimetry 2002) from January 2000 to December 2005. Envisat data are the ESA Radar Altimeter 2 (RA-2) GDR product available from September 2002 to October 2005 and described in ESA (2002).

Altimeter SWH measurements are extracted from the various products and selected according to the associated quality flags described in the user guides. Specific tests are performed for Jason-1 and Envisat based on the ratio of SWH standard deviation to SWH mean values established during the validation (Queffeulou 2004). These quality flags and tests, however, are not sufficient to discard all the erroneous SWH data. Spurious measurements are still observed: some are located in the vicinity of the coast where land can be within the altimeter footprint (some land data are also not flagged), or in oceanic regions of high scattering resulting in so-called sigma-zero blooms (Mitchum et al. 2004). Some other individual spurious measurements (corresponding mainly to high SWH) are not explained. Consequently, the data are filtered to eliminate these measurements. The screening is based on the analysis of the differences between successive along-track SWH measurements. For each pass (half-orbit) the mean value and standard deviation of differences of SWH measurements from pairs of consecutive points are estimated. At 1-s along-track sampling, two consecutive points are separated by about 6 or 7 km. A range is then defined by the mean value of the differences plus or minus 3 times the standard deviation. Individual data for which the differences with its neighboring measurements are outside this range are then discarded. Specific thresholds are also used at the beginning or at the end of continuous along-track series (i.e., corresponding to over land passes or to flagged data series). The entire dataset from each of the six altimeter missions was processed in this way. The number of discarded data points is low (a few measurements per pass, when it happens).

To compute long-term along-track statistics of TOPEX and Jason, new datasets are produced through interpolating the data along track to points separated by 0.05° of latitude. This corresponds more or less to the initial along-track 1-s time sampling of the altimeter.

The last step consists of correcting the altimeter SWH measurement. Comparisons with buoy data (Queffeulou 2004) show that the altimeter SWH estimate is generally in good agreement with the in situ data. Standard deviations of differences are on the order of 0.30 m. It is observed that the altimeter tends to slightly overestimate low SWH and to underestimate high SWH. Corrections to SWH were then established. These corrections, generally linear, correspond to a few percent of SWH. In addition, TOPEX SWH was also corrected for a drift between 1996 and 1999. For GFO, the correction was established for the global oceans through comparisons with TOPEX and ERS-2. The resulting corrected GFO SWH values demonstrate a few centimeters overestimated for the Mediterranean Sea relative to the other corrected altimeter data. As the GFO correction might not be suitable at low–medium SWH, it was decided to use uncorrected GFO data. All the applied corrections are provided in Table 6 of Queffeulou (2004).

3. Seasonal spatial features of SWH

The goal of this investigation is to retain the fine along-track sampling to study long-term characteristics of short-scale spatial variations of SWH over the Mediterranean Sea. The longest time series over the same ground track network is provided by TOPEX, from cycle 1 (25 September 1992) to 364 (11 August 2002), followed by Jason from cycle 22 (11 August 2002) to today (cycle 143, 3 December 2005). One question is raised by the time sampling: Is the 10-day repeat cycle sampling suitable to provide valuable long-term statistics? This can be illustrated using ERS-1 with the 3-day repeat cycle during the commissioning phase A and the two “ice” phases B and D. During phase A, a particular track of interest was identified crossing the Gulf of Lion (Fig. 2). Twenty-nine consecutive passes are available along this track, one every 3 days, covering a 3-month time period from 15 September to 15 December 1991.

Figure 3a shows the SWH data for each pass. The mean value of SWH over the 29 passes, calculated as a function of the latitude, is shown by the solid line in Fig. 3b. The initial dataset, sampled at 3 days, can be resampled at 9 days and 3 different series are considered, resulting from a 3-day shift in the initial series as shown by crosses, squares, and circles in Fig. 3a. Figure 3b shows that the along-track time average as a function of the latitude is significantly affected by the 9-day sampling case. For instance, the along-track mean SWH difference is about 75 cm (75%) between the 3-day sampling (solid line) and the circle line representing 9-day sampling. Differences between two 9-day samplings, as for instance between the square and circle lines (sampling shifted by 3 days), can be even larger. Indeed for the circle line case, the two largest SWH events over the 3 months (mistral on 20 October and on 7 November) are sampled, as shown in Fig. 3a.

This clearly illustrates that because of the time scale of the meteorological events in this particular region, a 9-day sampling is too long to get a correct estimate of the along-track mean value of SWH over a 3-month time period. A fortiori, the 10-day TOPEX and Jason sampling is not appropriate. Nevertheless, one can expect that over a very long time period, such as 13 yr, the seasonal meteorological events can be assumed to be statistically uniformly distributed relative to the altimeter sampling, so that the 10-day sampling does not affect the estimate of the mean value. An example of the along-track monthly mean value of SWH over 13 yr (for January 1993–2005) is shown in Fig. 4. SWH features are globally consistent. Some inconsistencies, however, may appear in some crossing point locations, as for instance at the crossing points in southeastern Sicily (between 15° and 20°E at about 35°N). This demonstrates that the 10-day sampling period is too long to properly get the monthly mean over 13 yr in this particular region. For this reason, only seasonal maps are presented.

The number of samples per altimeter cell (defined along track, interpolated every 0.05° of latitude) is about 9 over 3 months, and 117 for a season over the 13 yr. Some of the data are discarded by the filtering process. The analysis of the distribution of the number of samples per altimeter cell for the Mediterranean basin over the 13 yr of winter months (December–February) shows that the maximum of the distribution ranges between 108 and 115 samples. In practice the seasonal maps presented here are estimated using the altimeter cells with a sampling number larger than 70. This threshold was adjusted empirically. As expected the cells with the lowest data number are generally located at the coastal boundaries. Near the land-to-sea transition, the altimeter data can be invalidated either by the quality flags or by the data filtering process. Depending on the distance to the coast and the coastal topography, some altimeter data can go through the quality test though still being affected by the land presence in the footprint and thus resulting in higher values of SWH. In examining the maps of mean SWH, high values could occur near the coast with the number of such occurrences decreasing as the data number threshold increases. It is also noted that near the coast the data gap is generally less for the sea-to-land transition case than for the land-to-sea transition one. This is a result of the altimeter requiring some time to reacquire the sea surface after leaving land.

Using the monthly mean annual variation of SWH (section 4), the seasons are defined relative to the winter covering from December to February. Maps of seasonal mean values and standard deviations of SWH are shown in Figs. 5 and 6. The along-track SWH mean values of Fig. 5 show a rather good consistency between ascending and descending tracks. The maps reveal high variability over the seasons and over the various basins. Sea states resulting from local wind regimes can be clearly identified and characterized. The SWH maximum observed over the Gulf of Lion in all four seasons, with a relative minimum in summer, is induced by mistral and tramontane events (cold northwesterly to northerly winds). The extension toward the south during winter and autumn is well marked. Interesting details are shown on the zoomed portion of the winter map: sharp SWH gradients are observed near the southern coast of France, due to the fetch effect; the western boundary of the mistral region, from the eastern part of the Pyrénées in the northeast corner of Spain to Minorca, is well defined; some sheltering effect is also observed between the Mallorca and Minorca Islands (green along-track portion). The extension of relatively high SWH between Sardinia and the north coasts of Algeria and Tunisia during winter, spring, and autumn, and also in spring in the Strait of Sicily, is a result of the warm desert southeasterly to southwesterly wind regime (sirocco, called also marin in the Gulf of Lion). During winter, an SWH maximum is also observed over the western part of the eastern Mediterranean basin, south of Greece and Sicily. Mistrals can extend up to this region through the Strait of Sicily, but during winter the bora, a strong cold katabatic northerly wind originating from Yugoslavia, and the gregale (northeasterly) extend over the region.

In summer there is a strong contrast between western and eastern parts of the eastern Mediterranean basin. The eastern part (south of Crete and Rhodes and Cyprus) is characterized by a relatively high mean summer SWH (the maximum over the whole basin). This sea state is produced by the southerly extension, between Rhodes and Crete, of the northern monsoonal Aegean Sea etesian winds (meltem in Turkish), which have their maximum strength and occurrence in July and August. Note that very few measurements are available over the Aegean Sea due to both the size of the track network and the presence of many islands in the altimeter footprint polluting the measurement.

Over the Adriatic, SWH mean values are relatively low for all seasons, though bora katabatic winds are generally strong and produce high air–sea thermal contrast. Sea state development in this region is generally limited by the fetch.

Maps of seasonal SWH standard deviation (Fig. 6) indicate a maximum of variability over the Gulf of Lion region in all seasons. This is a result of the frequent occurrence and short duration (a few days in general) of mistral events. Though the mean SWH is high in the eastern most part of the Mediterranean during summer, the standard deviation is relatively low, thus suggesting some steadiness of the etesian extension wind regime during this season.

4. Annual and interannual variability

As seen above, one satellite like TOPEX with a 10-day repeat cycle is not able to fully describe the monthly SWH statistics. To increase the sampling, TOPEX and Jason data were augmented using data from all available altimeter missions (i.e., ERS-1, ERS-2, Geosat Follow-On, and Envisat). First, the monthly mean and standard deviation of SWH measurements were estimated over the whole Mediterranean basin for the 13 yr of available data. Then, the annual variability of the various subbasins was analyzed.

a. Global Mediterranean basin

Figure 7 shows the monthly results for each altimeter: the number of data values for each altimeter (Fig. 7a) and for the total of the missions (Fig. 7b), monthly mean values (Figs. 7b,c) with a zoom (Fig. 7e) and standard deviations of SWH (Fig. 7d). The number of data values depends on the phase of the missions, as described in section 2 and Table 1. The ERS-1 OPR data gaps are noticeable. The ERS-2 data number decreases drastically after June 2003 due to the onboard tape recorder failure. After this date the acquisition is restricted to areas where the satellite is in visibility of acquisition stations (i.e., mainly the northeastern Atlantic Ocean). Nevertheless, this gap can be filled using Envisat, which has the same orbit as ERS-2. The differences in data number between the satellites can also result from the varying sea coverage depending on the orbit. The total of all altimeter samples summed over the missions (Fig. 7b) clearly shows the advantage of multiple missions.

Figure 7c shows that the mean SWH values for the whole basin are generally in good agreement among the altimeters. Some discrepancies observed at the beginning or at the end of individual time series correspond to incomplete monthly observations (as for ERS-1 in December 1993, March 1995, or for Jason in January 2002). Figure 7e zooms in on SWH mean values from January 2000 to January 2006. GFO (green) and Jason (red stars) SWH values seem to be overestimated by a few centimeters in summer, a period of low SWH. The general agreement is good, however, and one can suggest that with minor corrections the agreement could be excellent. This tends to demonstrate that one satellite is able to provide monthly mean SWH variations for the whole Mediterranean basin. The SWH variability on the whole basin scale is very large, as shown in Fig. 7d. The monthly SWH standard deviation ranges between 0.3 and 1.2 m, which is almost comparable to the mean value fluctuations of 0.6–2.0 m.

As shown in the next section, the variability among the various subbasins is very large. Thus the time series of SWH mean values over the whole Mediterranean Sea has little geophysical meaning. Merging data from the six altimeters enables one to obtain meaningful estimates of the mean annual cycle. The mean SWH maximum is between 1.45 and 1.50 m observed in the December–February period. The summer minimum is also almost constant between 0.80 and 0.84 m. The interannual variability (Fig. 7c) is generally low for summer months, with the exception of summer 2003. For the years 1992 through 1996, some winter periods have values that are over the mean annual SWH, but the month-to-month variability is high. Winters 1997 through 2001 seem to be within the mean winter SWH, while winters 2002 through 2005 are clearly above the mean value. Note also in Fig. 7e the unusually sharp decrease of SWH as measured by all the altimeters for the periods between December 2001 and January 2002, and between February and March 2003.

These anomalies relative to the whole Mediterranean basin cannot be easily interpreted particularly because, as shown by the maps in the previous section, the seasonal characteristics are different for the various individual basins. The SWH variability in each subbasin is analyzed in the next subsection.

b. Subbasins

The particular subbasins (Fig. 1) corresponding to specific geographical topography and local meteorological characteristics are analyzed. Basic global statistics are given in Table 2 for the 14-yr time period. The maximum mean SWH (1.39 m) is observed in the Gulf of Lion region, which also has the largest standard deviation (1.07 m). A minimum of 0.85 m is observed in the Adriatic with a standard deviation of 0.60 m. Because of the varying sizes of the geographical areas, the number of data values differs significantly depending on the area.

SWH distributions are shown in Fig. 8. On the graphs the upper limit of SWH is restricted to 5 m because the statistics obtained for high SWH are not reliable due to the poor time sampling and the low occurrence of such events. The largest differences are observed between the Adriatic (black dot) and the Gulf of Lion (red circle). In the Adriatic, 80% of the data are less than 1.10 m. This same percentage yields a value of about 1.9 m for the Gulf of Lion. The maximum of the distribution’s spread is between 0.3 and 0.7 m for the Adriatic and between 0.6 and 0.9 m for the Gulf of Lion. The lowest SWH are observed increasingly in the Adriatic, Black Sea, Balearic, and Aegean basins, respectively. The largest SWH values, over 1 m, are observed in decreasing order in the Gulf of Lion, West, and Alboran regions (see Fig. 1).

1) Monthly annual cycle

The time sampling is too sparse to analyze the interannual variability on a monthly basis, but the mean monthly annual cycle can be estimated over the 14 yr. Before presenting the results for the various regions, a test is performed to compare the mean monthly annual cycles obtained from the altimeters over the Gulf of Lion region and from the meteorological buoy 61002 maintained in the Gulf of Lion (42.10°N, 4.70°E) by Météo France. The buoy data were provided by Météo France as part of the satellite validation activities of the Centre ERS d’Archivage et de Traitement (CERSAT), at Institut Français de Recherche pour l’Exploitation de la Mer (IFREMER). The SWH hourly buoy measurements cover a 3.3-yr time period from 4 December 2001 to 16 April 2005 with very few data gaps and a data return rate of 98%. Monthly values of the mean value and standard deviation of SWH were estimated from the 3.3-yr time period from the buoy measurements and from the altimeter data restricted to a limited area around the buoy location (41°–44°N, 3°–6.5°E) over the same time period. Mean annual cycles, normalized values, and scatter index (ratio of the monthly standard deviation to the monthly mean) are compared in Fig. 9. The agreement can be considered very good. The shapes of the annual cycles are the same, with a 10–20-cm bias at the highest SWH values. Accordingly, the altimeter results can be considered with some confidence concerning the annual cycle estimate.

Monthly annual cycles for the various regions are reported for the western (Figs. 10a,c,e) and eastern (Figs. 10b,d,f) basins: monthly mean values (Figs. 10a,b), monthly mean values normalized by the annual SWH mean (Figs. 10c,d), and scatter index (Figs. 10e,f). In the western part of the basin (Fig. 10a), the maximum amplitude of the mean annual cycle is observed in the Gulf of Lion with 0.93 m in August and 1.90 m in December. The Tyrrhenian and Balearic Seas have almost the same annual cycle: a minimum of about 0.7 m in June–August and a maximum of 1.3 m in December. The lowest values are observed in the Adriatic: 0.6 m in August and 1.15 m in December. Figure 10c shows the SWH normalized annual cycles. The annual cycle shapes are very similar for the Gulf of Lion, Tyrrhenian, Balearic, and Adriatic regions. In the western part of the basin, the Alboran region clearly demonstrates a specific annual cycle. The winter period is longer with SWH being near its maximum value from December to April (Fig. 10c, circles). The annual amplitude is low from 0.9 m in August to about 1.2 m in winter (Fig. 10a, circles). This area is known to exhibit long winter and spring seasons, comprising November through February and March through May, respectively, and “is noted for periods of stormy, winter-type weather alternating with a number of false starts of settled summer-type weather” (Brody and Nestor 1980).

In the eastern part of the basin (Fig. 10b), the maximum winter mean SWH values decrease from 1.7 m in the “West” region to 1.6 m in the “East” region, 1.4 m in the Aegean, and 1.3 in the Black Sea. For the summer minimum the lowest value is observed in the Black Sea (0.7 m) and the highest one in the East region (1 m). This is in contrast with 0.75 m observed in the West region and already noted in the summer map of Fig. 5. The annual cycle patterns are very similar during winter and spring (Fig. 10d). In summer, a significant difference exists between the West and East regions.

Comparisons of Figs. 10a–d show differences between the western and eastern parts of the Mediterranean basin. First, in the western side the winter SWH maximum occurs in December while it is observed in February in the eastern side (except in the West region where differences between December, January, and February are very low; star lines in Figs. 10b,c). Second, the decrease of SWH in the transition from winter to summer is faster in the eastern side than in the western side. Note also that in the eastern side the four subbasins have almost the same behavior between January and May, which is not the case for the second part of the year (Fig. 10d). Then, in the western side the summer minimum is roughly observed in August (low SWH, decreasing from May–August), while in the eastern side the summer minimum is observed in May–June (low SWH, increasing from May–June to August). The exception is in the West region for which the shape of the curve (Fig. 10d, star line) resembles more closely the western-side curves (Fig. 10c, star, cross, and solid lines) than the eastern-side ones.

The scatter index in Figs. 10e,f is the ratio of the monthly standard deviation to the monthly mean value of SWH. In the western side (Fig. 10e), it appears to be noisy and almost constant over the year, between 0.60 and 0.75, with the exception of the Alboran region (circle line) for which a minimum value is observed in summer. In the eastern part (Fig. 10f), a summer minimum is more strongly present, particularly for the Aegean and East regions (cross and circle lines) in July and August. This might be related to either some steadiness of the etesian during summer or to relatively high SWH in the East region (consequently decreasing the ratio).

2) Interannual seasonal variability

Because of the small size of the subbasins and the altimeter time sampling, it was decided to analyze the interannual variability seasonally and not on a monthly basis. Figures 11 and 12 present the 14-yr time series of the seasonal mean value of SWH for the regions of the western and eastern Mediterranean Sea, respectively. The mean annual cycle has been reported on each graph (dashed lines).

As the annual cycles were discussed in the preceding section, the interannual changes are investigated here. The curves of Figs. 11 and 12 show that interannual anomalies can be more or less correlated between the various regions. For each region a normalized anomaly is estimated as the difference between the seasonal time series (solid lines) and the mean seasonal annual cycle (dashed lines) divided by the mean seasonal annual value. Correlation coefficients are then computed between the normalized anomalies for the various regions and presented in Table 3. The anomalies can be considered as correlated when the coefficient is larger than 0.50 (30% of data), and as uncorrelated when the coefficient is less than 0.25 (39% of data).

The Alboran and Balearic regions are strongly correlated (77%) for both positive and negative anomalies (see winter anomalies during the years 1994 to 1996, 1999, or 2003 in Fig. 11). The correlations between the Gulf of Lion and the Alboran (46%) or the Balearic (55%) are less. For instance the positive winter anomalies observed in winter 1994 and 1995 in the Alboran and Balearic regions are not seen in the Gulf of Lion. Inversely, a large positive anomaly is observed in the Gulf of Lion in winter 2005 but not in the Alboran or Balearic regions. The Tyrrhenian and Adriatic regions are also well correlated (64%), though the winter 1999 anomaly in Tyrrhenian is not observed in the Adriatic region. The winter 2001 anomaly in the Tyrrhenian and Adriatic is not observed in the three other western regions. Only the 2003 winter anomaly is observed in all the regions, including those of the eastern Mediterranean.

In the eastern Mediterranean (Fig. 12), the West and East regions appear to be rather uncorrelated, with the exception of the winters of 1995, 1997, 1998, and 2003. In Figs. 11 and 12, the correlation seems to be higher between the West region and the Gulf of Lion (51%) and between the West and Adriatic regions (58%) than between the West and East regions (46%). The East region is uncorrelated to the western subbasins and is poorly correlated to other regions (about 40% with the Aegean and Black Seas). The Black Sea anomalies are well correlated with the Aegean ones (67%), and the coefficient surprisingly reaches 47% with the Adriatic.

5. Conclusions

Fourteen years of altimeter SWH measurements from six satellite missions have been analyzed over the Mediterranean Sea. One advantage of the altimeter is the fine along-track spatial resolution. The data of both TOPEX and Jason missions enable the construction of seasonal maps of along-track SWH mean values and standard deviations. These reveal short-scale and regional sea state features associated with the specific meteorological patterns of the various geographical basins. Time series of monthly SWH mean values and standard deviations from each satellite and over the whole Mediterranean Sea are in good agreement, thus showing interannual variability. But this information for the whole basin is of minor geophysical interest because of the high time and space variability over the various subbasins. The use of the six altimeter missions together enables the investigation of the monthly annual cycles at the short length scales of the subbasins. Significant differences are observed between the western and eastern parts of the Mediterranean Sea. The annual SWH cycle changes in both shape and amplitude according to the subbasin. Analysis of the seasonal interannual variability confirms the existence of some degree of independence between the subbasins. For instance, SWH interannual anomalies can be correlated for the whole subbasins, as for winters of 1999 and 2003, or can be uncorrelated, as in winter 1994 and 1995 for the Gulf of Lion and the Balearic or Alboran regions. Thanks to multisatellite missions and homogeneous corrections of the altimeter data, SWH time and space characteristics were able to be obtained at regional short scales. These results are independent of numerical wind and wave models. This method can be applied to any geographical region, which is a major implication of the study.

It has been shown that over a narrow region such as the Gulf of Lion region, the SWH mean annual cycle estimated from a buoy dataset over 3.3 yr compares relatively well to the result obtained from altimeter measurements over the same period. Buoy data are essential for many obvious reasons, such as validation of other measurement systems or wave spectrum estimates that cannot be obtained from altimeter data. To estimate accurate monthly mean cycles, however, the buoy data return must be high. Long time series of continuous wave buoy measurements are rare. Furthermore, buoy measurement deficiencies may occur, particularly in bad weather characterized by large SWH values. Gaps in the buoy time series obviously decrease the accuracy of these estimates. This illustrates the interest in long-term altimeter data, even over regions with buoy measurements.

In considering shorter time scales one can raise the question, How many altimeters are needed to estimate monthly SWH mean value equivalent to that obtained from a buoy? Some tests were performed using measurements of the Brittany buoy maintained by Météo France and the Met Office (U.K.) at 47.5°N, 8.5°W, west of the French Atlantic coasts. Monthly mean SWH were estimated over a 10-yr time period (1995–2006). Because of data gaps in the time series, several months (7.5%) contain less than 15 days of measurements, which is insufficient to estimate a monthly mean. Analysis of the altimeter data shows that 100 km is a minimum distance from which to collect altimeter data around the buoy location to get a significant number of altimeter passes. In this case the monthly number of satellite passes is between 8 and 14 using two satellites (ERS-2 and TOPEX), between 15 and 20 with three satellite (adding GFO), and between 30 and 45 with four satellites (adding Jason). Making the assumption that the distribution of the passes during one particular month is regular enough (which is in reality not the case, depending on the relative phasing and orbit characteristics of the satellites), a minimum of four satellites is required to estimate the monthly SWH average. Nevertheless, to increase the data sample the length of the radius of the selection area around the buoy could be extended to over 100 km. But a maximum value for this distance has to be set according to the regional meteorological and sea state spatial scales.

For the wave climate study, one drawback of altimeter data is the relatively short time series (14 yr) and the inhomogeneous and irregular time and space sampling. This is a strong limitation to the use of altimeter data to estimate the extreme wave height distributions, as investigated by Alves and Young (2003). Long time series are currently available from weather numerical models. However, reanalysis such as ERA-40 presents strong temporal heterogeneities (Caires and Sterl 2005a) due to the changes in the input wind field quality and the implementation of satellite data assimilation into the wave model. This is further impacted by the number of satellites (Skandrani et al. 2004) and by the quality of the assimilated data. The quality of wave model outputs has improved and may be considered as sufficient at the global scale but not at the regional short scale as in the Mediterranean Sea, for which significant corrections are needed (Medatlas Group 2004). Satellite wave measurements and wave modeling are complementary and the present work needs to be pursued in comparing satellite results to the data from wave atlases at the global scale (Sterl and Caires 2005) and for the Mediterranean Sea (Medatlas Group 2004).

Acknowledgments

We are indebted to CERSAT and Denis Croizé-Fillon for data management and computing. We thank K. Whitmer for correcting the English grammar and style. Part of the work was funded by the European projects EnviWave and MFSTEP.

REFERENCES

  • Abdalla, S., , Bidlot J-R. , , and Janssen P. , 2005: Assimilation of ERS and Envisat wave data at ECMWF. Proc. 2004 Envisat and ERS Symp., Publication SP-572, Salzburg, Austria, ESA, ESTEC, CD-ROM.

  • Alves, J-H. G. M., , and Young I. R. , 2003: On estimating extreme wave heights using combined Geosat, TOPEX/Poseidon and ERS-1 altimeter data. Appl. Ocean Res., 25 , 167186.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Aouf, L., , Lefèvre J-M. , , and Hauser D. , 2006: Assimilation of directional wave spectra in the wave model WAM: An impact study from synthetic observations in preparation for the SWIMSAT satellite mission. J. Atmos. Oceanic Technol., 23 , 448463.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • AVISO, 1996: AVISO user handbook: Merged TOPEX/POSEIDON products. 3.0 ed. CLS-CNES AVI-NT-02-101, 201 pp.

  • Bentamy, A., , Ayina H-L. , , Queffeulou P. , , Croize-Fillon D. , , and Kerbaol V. , 2007: Improved near real time surface wind resolution over the Mediterranean Sea. Ocean Sci., 3 , 259271.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Brody, L. R., , and Nestor M. J. R. , 1980: Handbook for forecasters in the Mediterranean, Part 2: Regional forecasting aids for the Mediterranean Basin. Naval Environmental Prediction Research Facility Tech. Rep. TR 80-10, 178 pp.

  • Caires, S., , and Sterl A. , 2005a: A new nonparametric method to correct model data: Application to significant wave height from the ERA-40 Re-Analysis. J. Atmos. Oceanic Technol., 22 , 443459.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Caires, S., , and Sterl A. , 2005b: 100-year return value estimates for ocean wind speed and significant wave height from the ERA-40 data. J. Climate, 18 , 10321048.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • CERSAT, 1996: Altimeter and microwave radiometer ERS products user manual, version 2.2. CERSAT, IFREMER C2-MUT-A-01-IF, 136 pp.

  • Chen, G., , Chapron B. , , Ezraty R. , , and Vandemark D. , 2002: A global view of swell and wind sea climate in the ocean by satellite altimeter and scatterometer. J. Atmos. Oceanic Technol., 19 , 18491859.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • ESA, 2002: Envisat RA-2/MWR product handbook. Product Document PO-TN-ESR-RA-0050, version 1.2, 185 pp.

  • Greenslade, D. J. M., , and Young I. R. , 2004: Background errors in a global wave model determined from altimeter data. J. Geophys. Res., 109 .C09007, doi:10.1029/2004JC002324.

    • Search Google Scholar
    • Export Citation
  • Greenslade, D. J. M., , and Young I. R. , 2005: The impact of altimeter sampling patterns on estimates of background errors in a global wave model. J. Atmos. Oceanic Technol., 22 , 18951917.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Medatlas Group, 2004: Wind and Wave Atlas of the Mediterranean Sea. WEAO Research Cell, 420 pp.

  • Mitchum, G. T., , Hancock D. W. III, , Hayne G. S. , , and Vandemark D. C. , 2004: Blooms of σ0 in the TOPEX radar altimeter data. J. Atmos. Oceanic Technol., 21 , 12321244.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Naval Oceanographic Office/NOAA Laboratory for Satellite Altimetry, 2002: GEOSAT follow-on GDR user’s handbook. NOAA/NESDIS/ORA:E/RA31, 20 pp.

  • Picot, N., , Case K. , , Desai S. , , and Vincent P. , 2003: AVISO and PODAAC user handbook: IGDR and GDR Jason products. AVISO SMM-MU-M5-OP-13184-CN, PODAAC JPL D-21352, 107 pp.

  • Queffeulou, P., 2004: Long-term validation of wave height measurements from altimeters. Mar. Geod., 27 , 495510.

  • Quilfen, Y., , Chapron B. , , Collard F. , , and Serre M. , 2004: Calibration validation of an altimeter wave period model and application to TOPEX Poseidon and Jason-1 altimeters. Mar. Geod., 27 , 535549.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sasaki, W., , Iwasaki S. I. , , Matsuura T. , , Iizuka S. , , and Watabe I. , 2005: Changes in wave climate off Hiratsuka, Japan, as affected by storm activity over the western North Pacific. J. Geophys. Res., 110 .C09008, doi:10.1029/2004JC002730.

    • Search Google Scholar
    • Export Citation
  • Skandrani, C., , Lefevre J-M. , , and Queffeulou P. , 2004: Impact of multisatellite altimeter data assimilation on wave analysis and forecast. Mar. Geod., 27 , 511533.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sterl, A., , and Caires S. , 2005: Climatology, variability and extrema of ocean waves: The web-based KNMI/ERA-40 wave atlas. Int. J. Climatol., 25 , 963977.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sotillo, M. G., , Ratsimandresy A. W. , , Carretero J. C. , , Bentamy A. , , Valero F. , , and Gonzales-Rouco F. , 2005: A high-resolution 44-year atmospheric hindcast for the Mediterranean Basin: Contribution to the regional improvement of global reanalysis. Climate Dyn., 25 , 219236.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wolf, J., , and Woolf D. K. , 2006: Waves and climate change in the north-east Atlantic. Geophys. Res. Lett., 33 .L06604, doi:10.1029/2005GL025113.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Map of the Mediterranean Sea subbasins.

Citation: Journal of Atmospheric and Oceanic Technology 24, 12; 10.1175/2007JTECH0507.1

Fig. 2.
Fig. 2.

ERS-1 ground tracks during the 3-day repeat cycle of phase A. The arrow indicates the particular track of interest.

Citation: Journal of Atmospheric and Oceanic Technology 24, 12; 10.1175/2007JTECH0507.1

Fig. 3.
Fig. 3.

(a) Individual ERS-1 SWH measurements every 3 days along the particular track of Fig. 1 from 15 Sep to 15 Dec 1991. (b) Three-month averages of ERS-1 SWH measurements along the particular track as a function of latitude for the initial 3-day (solid line) and 9-day samples (circles, squares, or crosses).

Citation: Journal of Atmospheric and Oceanic Technology 24, 12; 10.1175/2007JTECH0507.1

Fig. 4.
Fig. 4.

Thirteen-year along-track mean SWH for January.

Citation: Journal of Atmospheric and Oceanic Technology 24, 12; 10.1175/2007JTECH0507.1

Fig. 5.
Fig. 5.

(a)–(d) Maps of 13-yr along-track seasonal mean SWH and (e) winter map zoom.

Citation: Journal of Atmospheric and Oceanic Technology 24, 12; 10.1175/2007JTECH0507.1

Fig. 6.
Fig. 6.

Maps of 13-yr along-track seasonal SWH standard deviation.

Citation: Journal of Atmospheric and Oceanic Technology 24, 12; 10.1175/2007JTECH0507.1

Fig. 7.
Fig. 7.

Mediterranean basin time series of (a) number of data for each altimeter; (b) total of the missions; (c), (e) mean value; and (d) std dev of SWH measurements of the 6 altimeters over 14 yr on a monthly basis.

Citation: Journal of Atmospheric and Oceanic Technology 24, 12; 10.1175/2007JTECH0507.1

Fig. 8.
Fig. 8.

Mediterranean subbasins: SWH (a) distributions and (b) cumulative distributions from the 6 altimeters over 14 yr.

Citation: Journal of Atmospheric and Oceanic Technology 24, 12; 10.1175/2007JTECH0507.1

Fig. 9.
Fig. 9.

Comparisons of SWH annual cycles from buoy (crosses) and altimeter (circles) measurements. (a) Monthly mean value, (b) normalized mean value, and (c) scatter index.

Citation: Journal of Atmospheric and Oceanic Technology 24, 12; 10.1175/2007JTECH0507.1

Fig. 10.
Fig. 10.

Monthly annual cycle of SWH, over 14 yr, for (a), (c), (e) western and (b), (d), (f) eastern Mediterranean subbasins. (a), (b) Monthly absolute and (c), (d) normalized mean values. (e), (f) Monthly scatter index.

Citation: Journal of Atmospheric and Oceanic Technology 24, 12; 10.1175/2007JTECH0507.1

Fig. 11.
Fig. 11.

SWH seasonal (solid lines) time series and mean annual cycle (dashed lines) for the western Mediterranean subbasins.

Citation: Journal of Atmospheric and Oceanic Technology 24, 12; 10.1175/2007JTECH0507.1

Fig. 12.
Fig. 12.

SWH seasonal (solid lines) time series and mean annual cycle (dashed lines) for the eastern Mediterranean subbasins.

Citation: Journal of Atmospheric and Oceanic Technology 24, 12; 10.1175/2007JTECH0507.1

Table 1.

Time characteristics of the altimeter data.

Table 1.
Table 2.

Significant wave height statistics, 1992–2005, all altimeters.

Table 2.
Table 3.

Normalized interannual SWH anomalies: correlation coefficient (percent) between the Mediterranean subbasins.

Table 3.
Save