Publications

  • F. Giorgi, E. Coppola, F. Raffaele
    [Show abstract] [Hide abstract]
    ABSTRACT: We analyze trends of six daily precipitation-based and physically interconnected hydroclimatic indices in an ensemble of historical and 21st century climate projections under forcing from increasing greenhouse gas (GHG) concentrations (RCP8.5), along with gridded (land only) observations for the late decades of the 20th century. The indices include metrics of intensity (SDII) and extremes (R95) of precipitation, dry (DSL) and wet (WSL) spell length, the hydroclimatic intensity index HY-INT and a newly introduced index of precipitation area (PA). All the indices in both the 21st century and historical simulations provide a consistent picture of a predominant shift towards a hydroclimatic regime of more intense, shorter, less frequent and less widespread precipitation events in response to GHG-induced global warming. The trends are larger and more spatially consistent over tropical than extratropical regions, pointing to the importance of tropical convection in regulating this response, and show substantial regional spatial variability. Observed trends in the indices analyzed are qualitatively and consistently in line with the simulated ones, at least at the global and full tropical scale, further supporting the robustness of the identified prevailing hydroclimatic responses. The HY-INT, PA and R95 indices show the most consistent response to global warming, and thus offer the most promising tools for formal hydroclimatic model validation and detection/attribution studies. The physical mechanism underlying this response and some of the applications of our results are also discussed.
    Journal of Geophysical Research: Atmospheres. 10/2014;
  • Climatic Change 07/2014; 125(1):53-65. · 3.63 Impact Factor
  • Source
    04/2014;
  • Source
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The Po River is a crucial resource for the Italian economy, since 40% of the gross domestic product comes from this area. It is thus crucial to quantify the impact of climate change on this water resource in order to plan for future water use. In this paper a mini ensemble of 8 hydrological simulations is completed from 1960 to 2050 under the A1B emission scenario, by using the output of two regional climate models as input (REMO and RegCM) at two different resolutions (25km-10km and 25km-3km). The river discharge at the outlet point of the basin shows a change in the spring peak of the annual cycle, with a one month shift from May to April. This shift is entirely due to the change in snowmelt timing which drives most of the discharge during this period. Two other important changes are an increase of discharge in the wintertime and a decrease in the fall from September to November. The uncertainty associated with the winter change is larger compared to that in the fall. The spring shift and the fall decrease of discharge imply an extension of the hydrological dry season and thus an increase in water stress over the basin. The spatial distributions of the discharge changes are in agreement with what is observed at the outlet point and the uncertainty associated with these changes is proportional to the amplitude of the signal. The analysis of the changes in the anomaly distribution of discharge shows that both the increases and decreases in seasonal discharge are tied to the changes in the tails of the distribution, i.e. to the increase or decrease of extreme events.
    Science of The Total Environment 03/2014; · 3.16 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We intercompare three gridded observed daily rainfall datasets over Africa (FEWS (Famine Early Warning System), GPCP (Global Precipitation Climatology Project) and TRMM (Tropical Rainfall Measuring Mission)) in order to assess uncertainties in observation products towards the evaluation of the performance of a Regional Climate Model (RegCM3) in simulating daily precipitation characteristics over a domain encompassing the whole African continent. We find that different observation products exhibit substantial systematic differences in mean rainfall, but especially in higher order daily precipitation statistics, such as frequency of wet days, precipitation intensity and extremes as well as maximum length of wet and dry spells. For example, FEWS shows mostly higher frequency and lower intensity events than TRMM and GPCP. Thus, the different datasets provide quite different representations of daily precipitation behavior. As a result, although RegCM3 captures pretty well the monsoon rainband evolution and exhibits a representation of daily precipitation statistics within the range of the observations, it performs differently with respect to the various products. For instance, it simulates more intense but less frequent events over East and Southern Africa than in FEWS and vice versa compared to TRMM. We thus highlight the uncertainty in observations as a key factor preventing a rigorous and unambiguous evaluation of climate models over Africa. Improving the quality and consistency of observation products is thus paramount for a better understanding of the response of African climate to global warming. Copyright © 2012 Royal Meteorological Society
    International Journal of Climatology 06/2013; 33(7):1805-1817. · 3.40 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Regional climate models can be used to examine the impact of global warming at the regional level for the possible emission scenarios. According to IPCC AR4, a number of studies noticed large inter model differences leading to large uncertainties in the projected future changes in different aspects of monsoon rainfall. The aim of this study to understand the different response of similarly forced model RegCM4.3 with different convection scheme over the Ocean. In this study, RegCM4.3 is used to generate the climate scenarios for South Asia CORDEX domain using boundary forcing of global coupled climate models GFDL-ESM2M and MPI-ESM-MR in control runs and two emission scenarios RCP4.5 and RCP8.5 obtained from CMIP5 archive. Emanuel convective parameterization scheme has been used over the land in all the experiments. Over the Ocean, experiments are designed using both Emanual and Grell convection schemes. Rest of the model configuration is based on the different sensitivity experiments conducted to reduce the model temperature and precipitation biases. The model has been integrated at 50km resolution starting from the initial condition of 1st Jan 1970 and the climate simulation continued till 1st Jan 2100. Results indicate that all India JJAS mean temperature change is consistent with clear projection of warming, whereas the projected JJAS precipitation change shows large spread and uncertainty in trends. Even the similar forcing experiments with difference only in the convective scheme over ocean show large difference in their projected precipitation. Climatology of JJAS precipitation over Indian land and low level westerlies for the historical period (1979-2005) is well represented. However, there are differences in the precipitation distribution at regional level within the experiments mainly over the Northwest and southern part of Peninsular India. In case of Emanuel over ocean, enhanced convection over Indian land is mainly associated with the weak westerlies over the Bay of Bengal. The positive anomalies of westerly jet are captured over the Arabian Sea in both the cases where Grell convection over ocean is used in RegCM4.3. Grell scheme over ocean has significantly improved the pattern of westerly flow in RegCM4.3 simulations forced with the considered GCMs.
    04/2013;
  • [Show abstract] [Hide abstract]
    ABSTRACT: The hydroclimatic regime variability over the Amazon and La Plata Basin, the two main basins of South America (SA), are affect by local climate feedbacks and climate patterns associated with SST anomalies. The local and remote impacts of climate change over SA are evaluated using the regional model RegCM4 driven by HadGEM, GFDL and MPI global climate models (RegHad_CLM, RegGFDL, and RegMPI, respectively). All simulations used CLM (Community Land Model) land surface scheme and Emanuel convective scheme, with 50 km of horizontal resolution; except for one simulation driven by HadGEM that used the BATS (RegHad_BATS) as a land surface scheme and a mixed convection scheme (Grell over land and Emanuel over ocean). Climate changes were evaluated by comparing the future (2070-2098) with the present climate (1975-2004) during the rainy season (December to April). For the present day climate, the regional simulations reproduce reasonably well the precipitation annual cycle over Amazon and La Plata Basin. In particular the RegCM4 runs driven by MPI and GFDL are quite good over the Amazon and RegCM4 runs driven by HadGEM for La Plata Basin. For the future climate projections, all the global models show a precipitation decrease over the central Amazon Basin, northeast Brazil, and northern of SA (ranging from -5 to -25%) for the end of century, except the GFDL. Over the La Plata Basin, mainly over northern Argentina, and western of Amazon Basin simulations indicate an increase of precipitation (ranging from 25 to 50%). These patterns are intensified in the RegCM4-CLM simulations. The exceptions are the RegHad_BATS results that shows a general precipitation reduction over almost all Amazon basin. Over the La Plata Basin all regional models presented a positive coupling between sea surface temperature (SST) in Niño 3.4 region and precipitation. With exception of RegHad_CLM, all simulations show a negative SST/Precipitation regression over the central and northeastern Brazil. The soil moisture feedback on precipitation is evaluated too by mean of a statistical approach. The RegCM-CLM simulations have a common feature and they show a similar behavior when the future changes are investigated. The RegCM-BATS shows a different soil moisture feedback picture. Of course this can be explained by the differences in the two land-surface schemes and in the precipitation change signal that comes out from the two sets of simulations.
    04/2013;
  • [Show abstract] [Hide abstract]
    ABSTRACT: For the phase I CREMA experiment the regional climate model RegCM4 has been used to simulate the present and future time period over 5 domains across the word (Africa, South America, Central America, West Asia and Mediterranean) to contribute to CORDEX for a total of 34 simulations, spanning the period 1970-2100. Three GCMs have been used taken from the CMIP5 ensemble and they are HADGEM2ES, MPI-ESMMR and GFDL-ESM2M. The selection has been made according to the best RegCM4 performances when driven by those GCMs. For the present day period (1975-2205) the RCMs are evaluated and compared with the GCMs in term of ensemble mean biases, root mean square error and correlation by mean of Taylor diagrams for the two seasons DJF and JJA. For the temperature and precipitation biases the global and regional models are quite comparable. When the 5 domains are taken into account in the Taylor diagrams the individual regional model simulations show consistent improvements compared to the global model. This is confirmed by all the domains and by all the model configurations. For precipitation the situation is more complex and requires a detailed analysis for the different domains and model settings. The change signal has been also analyzed for the period 2070-2100 compared to the reference period 1975-2005. The ensemble mean change for the regional models show a stronger warming signal over Western Africa for DJF and Central and Southern Africa for JJA when compared with the global model mean change and a weaker warming signal over the Mediterranean domain for JJA. For precipitation, the regional ensemble mean change shows a stronger drying over the Western and Eastern Africa and north-east South America continent for the JJA season and an opposite wet signal for Northern Europe when compared with the global ensemble mean change. Also in the central part of India the regional ensemble shows a weaker monsoon for the future when compared to the global ensemble change that show and increase of precipitation during the JJA season. A stronger drying is also evident in Mexico and Central America from the regional simulations. For DJF the regional change signal shows a wetter Mexico and a wetter south Brazil and Argentina compared to the dryer signal coming from the GCMs and a dryer east Brazil. Some of these differences can be attributed to the land-surface feedback mechanism that are modulated by local processes and internal model physics like for example has been shown from a previous study in Western Africa (Mariotti et al. 2011).
    04/2013;
  • [Show abstract] [Hide abstract]
    ABSTRACT: This study is based on the RegCM4.3 simulated climate projections for the South Asia CORDEX domain. The lateral boundary forcing of GFDL-ESM2M is used in the control run and the emission scenario RCP8.5 obtained from the CMIP5 archive. In all the experiments, Emanuel convective parameterization scheme has been used over the land and Grell convection scheme over the ocean. The model has been integrated at 50km resolution starting from the initial condition of 1st January 1970 upto 1st January 2100. The model simulated surface air temperature and rainfall are validated for the current climate over the reference period 1975-2004. The future simulations are divided into four 30-years time slices of 2010-2039, 2040-2069 and 2070-2099 referred to as the near-future, mid-future and far-future periods respectively. Comparison of JJAS mean model projected rainfall for the different periods in the future interestingly reveals systematic changes over most of the Indian regions. The farther we go into the future the larger is the change from the reference period. Seasonal mean rainfall weakens over the central India, northern Indian Ocean and equatorial Indian Ocean. On the other hand, it strengthens more and more over the Arabian Sea and northern Bay of Bengal. A major fraction of the changes happen to come from the convective component of the rainfall, which is attributed to the fact that Asian Summer Monsoon is largely driven by the atmospheric convection. Nevertheless, non-convective precipitation component partly contributes to the enhancement of rainfall in the Arabian Sea and northern Bay of Bengal and to its weakening in the north Indian Ocean. Atmospheric circulation, surface evaporation, and large scale moisture convergence over the region are found to change along the same line and explain the changes observed with rainfall.
    04/2013;
  • Laura Mariotti, Erika Coppola, Filippo Giorgi
    [Show abstract] [Hide abstract]
    ABSTRACT: The Phase I CREMA experiment is the first contribution of the RegCM regional modelling community to the Coordinated Regional Downscaling Experiment (CORDEX). It consists of a series of scenario simulations over five CORDEX domains completed with the RegCM4 RCM driven by different CMIP5 GCMs. For the African domain three simulations have been completed from 1970 up to 2099, two have been driven by HADGEM2ES for RCP4.5 and RCP8.5 with one RegCM configuration and one driven by MPI-ESMMR for RCP8.5 with a different RegCM configuration. While the temperature changes are strongly driven by the global model, the precipitation change patterns of the regional model are different compared to the global models. According to Mariotti et al. 2011, it is confirmed that the local processes and internal model physics are key elements in determining the precipitation change signal simulated by the nested regional model in this large domain experiment, especially over equatorial and tropical regions. In particular over Western Africa, the Sahel region and Southern Africa, there is a reduction of the precipitation in the early season of the monsoon and an increase in the late part of the season for the future scenario (Seth et al. 2010).
    04/2013;
  • [Show abstract] [Hide abstract]
    ABSTRACT: The climate variability over the South America (SA) is mainly controlled by the El Niño-Southern Oscillation (ENSO) phenomenon. ENSO positive phase (El Niño; EN) is typically associated with below (above) normal precipitation over northern and northeastern (southeastern) of SA. The opposite signal is normally observed during the ENSO negative phase (La Niña; LN). In the context of CORDEX, the objective of this study is to investigate the ENSO precipitation signal over SA in the present and future climates. For this, we used an ensemble of RegCM4 projections driven by three global models (MPI-ESM-MR, GFDL-ESM2M and HadGEM2-ES). Basically, we investigated two scenarios (rcp4.5 and rcp8.5) during austral spring (September-October-November; SON) for present (P: 1975-2005), near future (N: 2020-2050), and far future (F: 2070-2098) climates. EN and LN years are defined according to the Oceanic Niño Index (ONI) that needs to be higher (lower) than a positive (negative) threshold for least 5 consecutive over-lapping seasons. For present climate, the number of EL and LN years simulated by HadGEM2-ES has a good agreement with the observations, while MPI-ESM-MR and GFDL-ESM2M underestimate the number of EN years. All global models project an increase (decrease) of EN (LN) years in the rcp8.5 far future climate scenario - except for GFDL-ESM2M which shows a slight decrease of EN years; on the other hand, for the near future climate there is a large discrepancy in this signal. For present climate, the composites indicate that all RegCM4 simulations capture the positive anomalies of rainfall over Southeastern of SA (SESA) associated with EN years. Individual RegCM4 simulations show some differences in the spatial pattern and intensity of rainfall anomalies during EN years compared with observations. The RegCM4 simulations ensemble for present climate shows a good representation of the observed anomalies. For near and future climates, all RegCM4 projections indicate an increase of SON precipitation over SESA and a decrease over the north-northeast of SA. However, the precipitation anomalies associated with EN years are weaker than that observed and simulated in present climate; this would be explained by EN events weaker in the future than in present climate as projected by global models.
    04/2013;
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The characteristics of tropical cyclones (TCs) over the Central America CORDEX domain is examined for present and future climate conditions using an ensemble of RegCM4 projections nested in two CMIP5 global models. The regional climate model is first tested in a 22 year (1982-2003) simulation with lateral and surface boundary forcing from the ERA-Interim reanalysis, showing a generally good performance in reproducing the observed TC climatology. Four scenario simulation (1970-2100, RCP8.5) are generated using two model configuration (CLM-EMAN and BATS-Grell) and two driving GCMs (HadGEM2-ES and MPI-ESM). The simulations employing the Grell scheme produced too low TC counts, while those using the Emanuel scheme reproduced the observed climatology, especially when driven by the MPI-ESM model. The simulation of TCs was thus sensitive to both the model convections scheme and forcing GCM. Comparison of future and present day TC climatologies indicates that the frequency of future TCs decreases over the tropical Atlantic and the eastern Pacific coastal areas while it increases over the Central Pacific and the northern part of the Atlantic. We also find a consistent increase in the frequency of intense TCs and a northward shift of TC tracks over the Atlantis. These findings are generally consistent with previous global and regional modeling studies and are suggestive of important effects of global warming on the characteristics of TCs in the tropical Atlantic and Pacific basins.
    Climatic Change 04/2013; 125(1):1677-. · 3.63 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this study we analyze a set of regional climate model (RCM) simulations applying dynamical downscaling of global climate model (GCM) simulations over the Mediterranean region. The RCM simulations are completed with the Regional Climate Model RegCM, version RegCM4.3, and cover one of the domains specified by the international initiative Coordinated Regional Downscaling Experiment (CORDEX). Two GCMs were selected from the Coupled Model Intercomparison Project Phase 5 (CMIP5) ensemble to provide the driving fields for the RegCM: HadGEM2-ES and MPI-ESM-MR. The simulations consist of an ensemble including multiple physics configurations and different "Reference Concentration Pathways" (RCP4.5 and RCP8.5). In total 13 simulations were carried out with 6 model physics configurations with varying convection and land surface schemes. The simulations were completed as part of the CORDEX REgCM4 hyper-Matrix (CREMA) experiment. The horizontal grid spacing of the RCM simulations is 50 km and the simulated period in all cases is 1970-2100. The Mediterranean region was divided into 6 subregions: Spain, Alps, Balkan, Turkey, Adriatic and full Mediterranean Basin. Mean temperature and precipitation changes are assessed from monthly through seasonal to annual scale mainly focusing on the period of 2070-2099 with respect to the 1976-2005 corresponding model reference period. We analyze changes in mean climate as well as climate variability and different indexes of extreme events. The results show a clear warming trend by the end of the XXIst century over all subregions, especially in the warm season, along with a corresponding drying. The change of the interannual variability of precipitation and temperature show a prevailing increase, especially over the Iberian peninsula. Changes in extremes are spatially coherent for temperature, but more variable for precipitation. Although these results are generally consistent with previous scenario runs, we also find that the model exhibits significant sensitivity to the physics configuration used, particularly in the summer season, while in the winter the GCM boundary forcing appears to be more dominant.
    04/2013;
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Future climate projections performed with the Regional Climate Model (RegCM4) are used to analyze the future changes on inter-annual variability of precipitation and temperature over Mexico and Central America. Two different global circulation models from the Couple Model Intercomparison Project phase 5 (CMPI5) are used as boundary conditions for two different RegCM4 configurations, which result in four different climate projections. Through a comparison of the precipitation annual cycles in reference period with future simulations, a shift in the annual cycle is found over Northwestern Mexico and Central America. During the rainy season (June to September), it is found an increase in the inter-annual variability of precipitation and temperature, together with a warming greater than 4°C in the mean seasonal temperature and a drying of more than 20%. An increased warming on the Eastern Pacific Ocean compared to the Tropical North Atlantic Ocean potentially generates a strengthened North Atlantic Subtropical High Pressure and also a stronger Caribbean Low Level Jet. This future ENSO-like state appears to be the mechanism driving the drying over the region
    04/2013;
  • Kevin Walsh, Filippo Giorgi, Erika Coppola
    [Show abstract] [Hide abstract]
    ABSTRACT: Regional climate model projections over the Mediterranean region are analysed for the presence of intense, warm-core lows that share some of the characteristics of tropical cyclones. The results indicate that the number of such systems decreases in a warmer world, particularly in winter. Comparison of the simulated numbers to changes in relevant climate diagnostics suggests that numbers decrease due to an increasingly hostile environment for storm formation, combined with a general poleward shift in the incidence of wintertime lows over western Europe.
    Climate Dynamics 03/2013; · 4.23 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The skill of a regional climate model (RegCM4) in capturing the mean patterns, interannual variability and extreme statistics of daily-scale temperature and precipitation events over Mexico is assessed through a comparison of observations and a 27-year long simulation driven by reanalyses of observations covering the Central America CORDEX domain. The analysis also includes the simulation of tropical cyclones. It is found that RegCM4 reproduces adequately the mean spatial patterns of seasonal precipitation and temperature, along with the associated interannual variability characteristics. The main model bias is an overestimation of precipitation in mountainous regions. The 5 and 95 percentiles of daily temperature, as well as the maximum dry spell length are realistically simulated. The simulated distribution of precipitation events as well as the 95 percentile of precipitation shows a wet bias in topographically complex regions. Based on a simple detection method, the model produces realistic tropical cyclone distributions even at its relatively coarse resolution (dx = 50 km), although the number of cyclone days is underestimated over the Pacific and somewhat overestimated over the Atlantic and Caribbean basins. Overall, it is assessed that the performance of RegCM4 over Mexico is of sufficient quality to study not only mean precipitation and temperature patterns, but also higher order climate statistics.
    Climate Dynamics 02/2013; · 4.23 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The interactions between land cover change in Central Africa and the Africa monsoon is investigated using a regional climate model simulation (2000–2007) under a Congo Basin deforestation scenario. Decreased evaporation over the deforested area locally produces a heat low and reduced precipitation. In JJA, this low strengthens the West Africa monsoon causing increased precipitation over the Sahel and decreased precipitation over the Guinea coast. In DJF, it strengthens the south-equatorial African monsoon causing an increase of precipitation over south-equatorial Africa. Therefore our simulations indicate that the biosphere-state in Central Africa may play an important role locally but also remotely via interactions with regional monsoon dynamics. Copyright © 2012 Royal Meteorological Society
    Atmospheric Science Letters 01/2013; 14(1):45-51. · 1.75 Impact Factor
  • Climate Dynamics 01/2013; · 4.23 Impact Factor
  • E. Coppola, F. Giorgi
    [Show abstract] [Hide abstract]
    ABSTRACT: The RegT-band version of RegCM4.1 release is used in a surrogate climate change experiment. Ten years of RegT-band simulation have been completed from 1998 to 2007, using the ERA-interim as boundary condition. The same ten years have been completed with a modified version of the model with doubled greenhouse gas concentration and with an increase of 3 temperature degrees in the lateral and boundary conditions for both the atmosphere and the ocean. Change in the hydrological cycle are investigated and in particular the changes in the hydroclimatic intensity are presented by mean of the analysis of the HY-INT index (Giorgi et al., 2011). By using this index both precipitation intensity and dry spell length are taken into account in the same time and the increase of hydroclimatic intensity is confirmed according to the Giorgi et al., 2011 results.
    04/2012;

36 Following View all

66 Followers View all