«Atmosphere 2014, 5, 370-398; doi:10.3390/atmos5020370 OPEN ACCESS atmosphere ISSN 2073-4433 Article Patterns of ...»
Atmosphere 2014, 5, 370-398; doi:10.3390/atmos5020370
Patterns of Precipitation and Convection Occurrence over the
Mediterranean Basin Derived from a Decade of Microwave
Bahjat Alhammoud 1,2, Chantal Claud 1,*, Beatriz M. Funatsu 3, Karine Béranger 2
and Jean-Pierre Chaboureau 4
Laboratoire de Météorologie Dynamique-IPSL, CNRS, Ecole Polytechnique, 91128 Palaiseau,
France; E-Mail: firstname.lastname@example.org Department of Mechanical Engineering, ENSTA-ParisTech, 91762 Palaiseau, France;
E-Mail: email@example.com CNRS, LETG-Rennes-COSTEL UMR 6554, Université Rennes 2, 35043 Rennes, France;
E-Mail: firstname.lastname@example.org Laboratoire d’Aérologie, University of Toulouse and CNRS, F-31400 Toulouse, France;
E-Mail: Jean-Pierre.Chaboureau@aero.obs-mip.fr * Author to whom correspondence should be addressed; E-Mail: email@example.com;
Tel.: +33-169-335-129; Fax: +33-169-335-211.
Received: 13 January 2014; in revised form: 8 April 2014 / Accepted: 22 April 2014 / Published: 30 May 2014 Abstract: The Mediterranean region is characterized by its vulnerability to changes in the water cycle, with the impact of global warming on the water resources being one of the major concerns in social, economical and scientific ambits. Even if precipitation is the best-known term of the Mediterranean water budget, large uncertainties remain due to the lack of suitable offshore observational data. In this study, we use the data provided by the Advanced Microwave Sounding Unit-B (AMSU-B) on board NOAA satellites to detect and analyze precipitating and convective events over the last decade at spatial resolution of
0.2° latitude × 0.2° longitude. AMSU-B observation shows that rain occurrence is widespread over the Mediterranean in wintertime while reduced in the eastern part of the basin in summer. Both precipitation and convection occurrences display a weak diurnal cycle over sea. In addition, convection occurrences, which are essentially located over land during summertime, shift to mostly over the sea during autumn with maxima in the Ionian sub-basin and the Adriatic Sea. Precipitation occurrence is also inferred over the sea from two other widely used climatological datasets, HOAPS (Hamburg Ocean Atmosphere Atmosphere 2014, 5 371 Parameters and Fluxes from Satellite Data) and the European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis interim (ERA-Interim). There is generally a rather fair agreement between these climatologies for describing the large-scale patterns such as the strong latitudinal gradient of rain and eastward rain signal propagation. Furthermore, the higher spatial resolution of AMSU-B measurements (16 km at nadir) gives access to mesoscale details in the region (e.g., coastal areas). AMSU-B measurements show less rain occurrences than HOAPS during wintertime, thereby suggesting that some of the thresholds used in our method might be too stringent during this season. We also observed that convection occurrences in ERA-Interim are systematically lower than those derived from AMSU-B. These results are potentially valuable to evaluate the rainfall parameterization in weather and climate models and to constrain ocean models.
Keywords: precipitation; convection; microwave remote sensing; climatology; Mediterranean
1. Introduction The Mediterranean region (Figure 1) is a densely populated area and one of the most responsive to climate change (e.g., [1,2]). On the one hand, water availability is of great importance, and on the other hand, severe weather events, which cause heavy rainfall and floods, add to the socio-economical distress. In fact, trend analysis of daily rainfall data for the period of 1951 to 1995 based on 265 Mediterranean stations have shown that the Central–Western Mediterranean faced a change in the rainfall distribution with reduction of mean precipitation totals, as well as the whole Mediterranean Sea , but increase in the occurrence of heavy precipitating events (e.g., ). The expected increase in weather systems severity in the future would contribute to the increase in extreme rainfall .
Since the beginning of the 2000s, several studies have been carried out to investigate the precipitation climatology over the Mediterranean region. These studies have shown that rain gauges, satellite observation or reanalyses show reasonable agreement regarding large-scale rainfall distribution, while results differ in regional scale particularly over coastal areas and regions with complex topographies (e.g., [6,7]).
In general, the monitoring of precipitation at a high spatial and temporal resolution over the Mediterranean Sea remains challenging (see e.g., ). Conventional instrumentation for measuring precipitation such as rain gauges quantify the flux of rainfall and provide rain rate using direct measurements in a small area. Errors in the rain gauges measurements arise from several factors, such as short rainfall events, wind flow, evaporation and the spatial and temporal distribution of raindrop size [8–10]. In contrast, weather radar and satellites take a snapshot of the rainfall distribution over a limited area of about a 200 km radius around the site. In spite of uncertainty in the retrieval relationship reflectivity/rainfall from weather radars, they have gained a greater interest due to their capability to monitor rainfall at high spatial and temporal resolution.
Rain gauges and weather radar systems networks are unevenly distributed and mostly concentrated over land masses. Over the oceans, the few existing systems are mainly located on islands and are therefore subject to local influences. While both rain gauges and radars provide quite accurate Atmosphere 2014, 5 372 estimates of rainfall rate, the level of representation of the measurement relative to the surrounding region remains a major issue. The measurement of precipitation on a global basis must therefore rely on satellite systems that are able to provide information at regular intervals (for a review, see [8–11]).
Figure 1. Orography of the studied area (in meters).
wMed: The western Mediterranean basin consists of the Alboran (ALB), the Algerian (ALG), the Balearic (BAL), the Ligurian (LIG) and the Tyrrhenian (TYR) sub-basins; cMed: the central Mediterranean basin consists of the Adriatic Sea (ADR) and the Ionian sub-basin (ION); and eMed: the eastern Mediterranean basin consists of the Aegean Sea (AEG) and the Levantine sub-basin (LEV); BLK: Black Sea; ALP: Alps; ATL: Atlas; BAL: Balkan; PYR: Pyrenees; ANT:
Anatolian and TRS: Taurus Mountains. Blue-solid lines indicate limits between western, central and eastern Mediterranean basins.
It is generally admitted that passive microwave (PMW) techniques (e.g., the Special Sensor Microwave Imager (SSM/I), the Tropical Rainfall Measuring Mission (TRMM), the Advanced Microwave Scanning Radiometer—Earth Observing System (AMSR-E)) provide better instantaneous estimates of precipitation than visible/infrared techniques (e.g., [12–14]). The infra-red (IR)-based technique offers a very high temporal sampling (e.g., 15 min for GEO satellites) and a fine spatial resolution (down to 3 km). However, a limitation of the IR-based methods is that the relationship of surface rainfall and cold cloud tops is indirect and highly dependent on season and location. The problem is even more severe when dealing with multilayer cloud systems and the bad collocation of the coldest cloud tops and the heaviest precipitation. These issues limit the IR-algorithm’s applicability in global climate studies [9,14].
Atmosphere 2014, 5 373 PMW techniques exploit the fact that microwave radiation emitted from the earth’s surface interacts with the atmosphere constituents such as water vapor and cloud distribution and precipitation particles.
Therefore, the desired measurement (i.e., the precipitation signal) is mixed with the earth’s surface radiation. That means that depending on the surface emissivity, the measured signal can be affected.
For example, an over-ocean signal due to emission by rain subsequently results in a warmer area than the colder ocean background. Over land, the rainfall is detected by the reduction of the upwelling radiation due to scattering by large, frozen hydrometors. Overall, PMW retrieval methods work better over ocean than over land areas where the emissivities of soil and vegetation are greatly variable from place to place . The spatial resolution (5–60 km) of PMW sensor is constrained by the antenna size, but it still has adequate resolution for global and regional (such as the Mediterranean basin) climatological studies.
PMW captures the precipitation characteristics and provides a better measure of precipitation than IR, even if IR tracks faithfully the movement of precipitating system . Since polar-orbiting satellites that carry microwave sounders overpass the same point on the globe roughly twice daily, infrared observations from geostationary satellites have been consistently used for the characterization of diurnal aspects of precipitation. However, with the launch of a series of Advanced Microwave Sounding Unit (AMSU) instruments on board several platforms, the number of overpasses has increased significantly, providing an unprecedented opportunity to explore microwave-based data to characterize precipitation and convection over the Mediterranean region in a diurnal timescale.
In this paper, we use the data provided by the AMSU-B radiometer (replaced on recent platforms by the Microwave Humidity Sounder, or, MHS), which permits a concomitant observation of convection/precipitation over land and sea at a temporal resolution higher than that of meteorological analyses and with a spatial resolution of roughly 20 km. Brightness temperatures measured since 1999 by channels of the AMSU-B and MHS sensors in the water vapor absorption line (183–191 GHz) allow a screening of precipitation over the whole Mediterranean region, including the sea, where in situ observations are scarce, providing a complementary and completely independent picture from studies based on reanalysis and ground station data.
The present work aims at better characterizing the occurrence of precipitation and convection over the Mediterranean basin at a regional scale over the last decade. It is the first time that the monthly climatology of the MR and DC over the Mediterranean region with a spatial resolution of
0.2° × 0.2° latitudes is discussed. Compared to a number of climatic studies that have been conducted over this area, the spatial resolution is improved, and therefore we can expect more detailed results especially over the sea, where the monitoring remains challenging. It is a contribution to the HyMeX (Hydrological cycle in the Mediterranean eXperiment ) program whose objectives are to better understand and quantify the hydrological cycle and related processes in the Mediterranean, with an emphasis on (i) high-impact weather events, (ii) interannual to decadal variability of the Mediterranean coupled system, and (iii) associated trends in the context of global change. Climatological features of precipitation at regional scale will be presented with an emphasis on the Mediterranean Sea, separating moderate rain (MR) and deep convection (DC) occurrences following Funatsu et al. . Comparisons with the widely used HOAPS (Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data) observational dataset with a spatial resolution of 0.5° × 0.5° latitude–longitude, and with the European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis interim (ERA-Interim; resolution Atmosphere 2014, 5 374 of 0.75° × 0.75° latitude–longitude) will also be discussed in order to illustrate the advantages and shortcomings of AMSU-B products.
The paper is organized as follows. Section 2 describes microwave AMSU-B satellite observations and diagnostics used to detect precipitation and convection occurrences, and the other precipitation datasets considered in this paper. A general description of the monthly climatology of MR and DC occurrences is done in Section 3. Comparisons with other climatologies (HOAPS, ERA-Interim) at basin and regional scales are presented in Section 4. Finally, in Section 5, the results of the study are discussed and summarized.
2. Precipitation Products
2.1. AMSU-Based Precipitation and Convection Occurrence Since this dataset has been described in details in a recent paper , we recall here only the most important points. This study mainly uses the AMSU-B sensor, which is a cross-scanning microwave instrument flying on board NOAA sun-synchronous polar-orbiting satellites since late 1998. AMSU-B has three channels (3 to 5) centered around the water vapor absorption line (183 GHz) and has 90 fields of views (FOVs) along a wide (~2300 km) swath. FOV of AMSU-B ranges between 16 km at nadir and 36 km for the largest viewing angles . This sounder was initially designed for optimal moisture retrieval, but a number of studies have shown its utility for rainfall detection (e.g., [19,20]).
More details can be found in the NOAA’s KLM User’s Guide (http://www2.ncdc.noaa.gov/docs/klm/).
NOAA-15, 16, and 17 observations are considered in this study. The longest time series are provided by NOAA-15, which has been operational from January 1999 until July 2010. However, we do not use NOAA-15 observations after December 2007, due to scan asymmetry issues especially for channel 4 , which may affect deep convection determination (see below). We also dropped NOAA-16 measurement after December 2007 due to a residual uncorrected radio-frequency interference (in spite of initial corrections made by NOAA) and asymmetry problem .