Health & Lifestyle

A Statistical Adjustment of Regional Climate Model Outputs to Local Scales: Application to Platja de Palma, Spain

Description
1FEBRUARY 2012 A M E N GUAL ET AL. 939 A Statistical Adjustment of Regional Climate Model Outputs to Local Scales: Application to Platja de Palma, Spain A. AMENGUAL Grup de Meteorologia, Departament de
Published
of 14
25
Published
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Transcript
1FEBRUARY 2012 A M E N GUAL ET AL. 939 A Statistical Adjustment of Regional Climate Model Outputs to Local Scales: Application to Platja de Palma, Spain A. AMENGUAL Grup de Meteorologia, Departament de Física, Universitat de les Illes Balears, and Departament de Recerca en Canvi Global, Institut Mediterrani d Estudis Avancxats, Palma de Mallorca, Spain V. HOMAR AND R. ROMERO Grup de Meteorologia, Departament de Física, Universitat de les Illes Balears, Palma de Mallorca, Spain S. ALONSO Grup de Meteorologia, Departament de Física, Universitat de les Illes Balears, and Departament de Recerca en Canvi Global, Institut Mediterrani d Estudis Avancxats, Palma de Mallorca, Spain C. RAMIS Grup de Meteorologia, Departament de Física, Universitat de les Illes Balears, Palma de Mallorca, Spain (Manuscript received 21 December 2010, in final form 29 July 2011) ABSTRACT Projections of climate change effects for the System of Platja de Palma (SPdP) are derived using a novel statistical technique. Socioeconomic activities developed in this settlement are very closely linked to its climate. Any planning for socioeconomic opportunities in the mid- and long term must take into account the possible effects of climate change. To this aim, daily observed series of minimum and maximum temperatures, precipitation, relative humidity, cloud cover, and wind speed have been analyzed. For the climate projections, daily data generated by an ensemble of regional climate models (RCMs) have been used. To properly use RCM data at local scale, a quantile quantile adjustment has been applied to the simulated regional projections. The method is based on detecting changes in the cumulative distribution functions between the recent past and successive time slices of the simulated climate and applying these, after calibration, to the recent past (observed) series. Results show an overall improvement in reproducing the present climate baseline when using calibrated series instead of raw RCM outputs, although the correction does not result in such clear improvement when dealing with very extreme rainfalls. Next, the corrected series are analyzed to quantify the climate change signal. An increase of the annual means for temperatures together with a decrease for the remaining variables is projected throughout the twenty-first century. Increases in weak and intense daily rainfalls and in high extremes for daily maximum temperature can also be expected. With this information at hand, the experts planning the future of SPdP can respond more effectively to the problem of local adaptation to climate change. 1. Introduction Observations show that the global mean surface temperature has increased notably during the twentieth century. In fact, the second half of the twentieth century Corresponding author address: Arnau Amengual, Dept. de Física, Universitat de les Illes Balears, Edifici Mateu Orfila, Cra. de Valldemossa, km. 7.5, Palma de Mallorca, Spain. has been the warmest period for at least the last 1300 yr in the Northern Hemisphere (Solomon et al. 2007). According to the National Oceanic and Atmospheric Administration (NOAA) s National Climatic Data Center (Smith et al. 2008), the 14 warmest years in the instrumental record have been observed during the period. The rate of global surface warming from 1979 to 2005 is estimated at 0.278C decade 21. Furthermore, the estimated trends at regional scale for this interval show high spatial variability and, for the DOI: /JCLI-D Ó 2012 American Meteorological Society 940 J O U R N A L O F C L I M A T E VOLUME 25 Mediterranean area as a whole, the Intergovernmental Panel on Climate Change (IPCC) shows this trend is between 0.25 and 0.358Cdecade 21 (Solomon et al. 2007). Associated with this global warming, a redistribution of the rainfall and other atmospheric variables (e.g., pressure, wind, cloudiness) has been observed with even higher spatial variability than for temperature. For the whole Mediterranean region, observations indicate a decrease in the precipitation amounts estimated between 5% and 20% from 1901 to This reduction has appeared with decreases below 3% during the period (Solomon et al. 2007). However, the robust detection of precipitation trends is always problematic owing to their high spatial and temporal variability (Huntington 2006). The study of climate change from instrumental records aggregated in databases of regional range for example, the Mediterranean Basin can inevitably mask local features. Being aware of this problem, Homar et al. (2010) carried out a study of the recent trends in rainfall and temperatures for the Balearic Islands. The results of their study indicate a decrease in annual rainfall amounts at a rate of 16.6 mm decade 21 for the second half of the twentieth century. Autumn and winter decreases are responsible for most of the decline in precipitation. The extreme of daily precipitation that is, either the smallest or largest values are becoming more frequent together with a decreasing contribution of the intermediate amounts to the total accumulations. Furthermore, the study reveals that minimum and maximum temperatures over the last decades of the twentieth century have risen at a rate of and 0.488C decade 21, respectively. This regional warming is most evident for springs and summers. The Consortium of Platja de Palma an agreement signed by the Balearic Islands government and the Ministry of Industry, Tourism and Trade of the Spanish government for the redesign and suitability to the needs of the twenty-first century of this important tourist resort was set up with the aim of restructuring the tourist industry for the System of Platja de Palma (SPdP). This emplacement is one of the major resorts in the whole Mediterranean region, and it is entirely devoted to a mass tourism model (Fig. 1; further information at consorcioplayadepalma.es). The consortium has geared its work toward configuring a different vision and contemplating a new tourist model suitable to the needs of the twenty-first century. Major guidelines for the consortium include as key issues sustainability, climate and global change, and social and residential cohesion. Therefore, the assessment of the consequences of climate change and the subsequent implementation of adaptation strategies has become one of its main concerns. Within this framework, we explore future climate projections, and specifically, the projected changes of the most relevant meteorological parameters for carrying out suitable sun, sea, and sand (3S) leisure activities. Specifically, we explore the shifts in the future annual and seasonal mean regimes as well as the changes in the extreme climatic events. Both topics have a high social and economic interest, since the productive enterprise developed in tourist regions are strongly linked to its climatology, being extremely sensitive to extreme weather impacts as well (Amelung and Viner 2006). Atmosphere ocean general circulation models (AOGCMs) constitute the primary tool to produce future climate projections. AOGCM simulations have been run under a wide range of scenarios for greenhouse gas emissions and aerosols [Special Report on Emissions Scenarios (SRES); Nakicenovic et al. 2000). These scenarios describe plausible evolutions for these emissions depending on socioeconomic conditions and world development guidelines. The A1B scenario has widely been adopted in the latest climatic simulations, since it is halfway between the most unfavorable (A2) and optimistic (B1) scenarios (see Action5htmlpage&page5welcome). 1 Although these models are suitable to provide future global climate scenarios, their coarse spatial resolutions are not appropriate to evaluate most regional and local impact studies. Although climate change is a problem of global causes and consequences, its impacts become apparent locally. Climate change analysis at regional and local scales, such as an increase in the frequency and/or intensity of extreme events, requires quantitative estimations at increased spatial and temporal resolutions. Dynamical downscaling applied to AOGCM outputs attempts to account for the effects of mesoscale forcings and other subgrid-scale features by nesting increasedresolution regional climate models (RCMs) in large-scale meteorological fields generated by general models (Giorgi and Mearns 1999; Denis et al. 2002; Beck et al. 2004). For the western Mediterranean area, regional effects exert a particularly strong influence on the distribution of meteorological variables owing to the characteristic configuration of land and sea and complex orography (Fig. 1; Amengual et al. 2007). Although regional climate modeling often improves the performance from AOGCMs at regional scales, the resolution still remains inadequate to address uncertainties arising from several sources. We present a new statistical approach based on the application of a quantile quantile (Q Q) adjustment to several projections of RCMs to better assess climate change impacts over the SPdP. Simulated daily data have been 1 Note that the global climate modeling community is currently using the representative concentration pathway (RCP) scenarios. 1FEBRUARY 2012 A M E N GUAL ET AL. 941 FIG. 1. Geographical location for the SPdP in the western Mediterranean region. Major topographic features for the entire area and Mallorca Island are shown. Also displayed is the location of the automatic weather station (LEPA). provided by a set of regional climate models run under the A1B emissions scenario. The rest of the paper is structured as follows: section 2 contains a brief description of the study area, addressing the main environmental, social, and economic issues; section 3 describes the observed and simulated databases that have been used, as well as the proposed empirical correction method and it presents the results of a validation test applied to this adjustment technique; section 4 discusses the projected annual and seasonal variations of the mean and extreme regimes for the parameters of interest; and finally, section 5 provides an assessment of the implemented approach and its later applications. 2. Overview of the study area: Climatic characteristics and economic activities The climate of the Balearic Islands is characteristic of the western Mediterranean region. It is associated with a wide range of synoptic flows and is strongly influenced by the Mediterranean Sea, which is the main source of moisture for the region. In summertime, the Azores s high pressure system dominates the synoptic situation, producing a sustained increase of the air and sea surface temperatures together with a period of reduced precipitation. When the eastward extension of the Azores high pressure system moves equatorward in early autumn, it facilitates the arrival of mid- to upper-level Atlantic cold air masses. Given the presence of moist and warm Mediterranean air at low levels in autumn, relatively high sea surface temperatures, and the complex orography of the region, this is when heavy rainfalls are most common (Tudurí and Ramis 1997; Romero et al. 1999; Amengual et al. 2008). The northern side of Mallorca can be affected by intense northerly to westerly flows linked to the passage of midlatitude fronts, mostly in winter. This area also has the largest number of days in Mallorca with strong winds, cloudiness, and rain (the mean annual precipitation amounts are about mm). In contrast, moist easterly flows produce most of the rainy days (typically convective in character) in the southern part of the island, which can result in very high precipitation rates, principally in autumn, but with occasional events in spring and winter (Romero et al. 1999). Furthermore, Mallorca constitutes a perfect example of a region with complex topography. The island is characterized by an irregular distribution of mountainous ranges 942 J O U R N A L O F C L I M A T E VOLUME 25 and plains. The highest elevations are in the northwestern mountainous region the Tramuntana range with elevations close to 1500 m (Fig. 1). Mallorca s rugged relief together with an irregular coastline and its inherent insular characteristics (e.g., evident in the summer seabreeze regimes) result in large climatological diversity (Guijarro 1986). For example, mean annual rainfall amounts in the Serra de Tramuntana can range from 1000 to 1500 mm, whereas in the southern coastline only 50 km away the amounts barely exceed 350 mm. For the Balearics, the mean annual rainfall amount is roughly 560 mm (Homar et al. 2010). Also, temperatures strongly depend on elevation and distance from the sea. Likewise, the western Mediterranean area is quite often affected by the incursion of Siberian and Saharan and occasionally Arctic air masses, resulting in severe cold and heat waves. Mean annual minimum and maximum temperatures in the Balearic Islands are and 21.88C, respectively (Homar et al. 2010). The SPdP is located in the southwestern coast of Mallorca, neighboring the city of Palma (Fig. 1). It is the main tourist region in the Balearics and is situated within the municipalities of Palma and Llucmajor. The average of nights per year spent by visitors in the entire Palma county was more than 8 million during the period (taking into account hotels and tourist apartments; INE 2010). In particular, the SPdP can host visitors per night with a total amount of 1.2 million nights spent per year. Its main socioeconomic activities rely on beachbased tourism, although it is also an important residential area permanent inhabitants owing to be a dormitory town of the main city of the Balearics (Fig. 1). It has 10-km coastline, mainly consisting of sand beaches, and the dominant holiday activities are those related to the sun, sea, and sand tourism mass model. 3. Database and methods a. Input data and quantile quantile adjustment Observations were obtained from the automatic weather station deployed in 1973 by the Spanish Meteorological Agency [Agencia Estatal de Meteorología (AEMET)] at Palma s international airport (denoted as LEPA; N, 2.438E). No significant construction has been undertaken near this station, located at the head of the first runaway, far from any urban development (Fig. 1). Therefore, local effects from urbanization, such as heat island warming or precipitation sheltering, are safely negligible (Gual et al. 2002). Besides, its proximity to the SPdP less than 4 km makes it an ideal dataset to address the objectives of this work. To characterize the evolution of the meteorological variables over the SPdP, TABLE 1. List of transient RCM experiments driven within the ENSEMBLES European project for the period. Note that all the models have a spatial resolution of 25 km and have been run under the SRES A1B. Driving GCM RCM Acronym Institute ECHAM5 RCA3 C4IRCA3 C4I ARPEGE HIRLAM DMI-HIRLAM5 DMI ECHAM5 HIRLAM DMI-HIRLAM5 DMI BCM HIRLAM DMI-HIRLAM5 DMI HadCM3 CLM ETHZ-CLM ETHZ ECHAM5 RegCM ICTP-REGCM ICTP ECHAM5 RACMO KNMI-RACMO KNMI HadCM3 HadRM3Q0 METO-HC-HadCM3Q0 HC HadCM3 HadRM3Q3 METO-HC-HadCM3Q3 HC HadCM3 HadRM3Q16 METO-HC-HadCM3Q16 HC BCM RCA SMIRCA SMHI ECHAM5 RCA SMIRCA SMHI HadCM3 RCA SMIRCA SMHI complete daily series of 2-m minimum and maximum temperatures, accumulated precipitation, 2-m mean relative humidity, mean cloud cover, and 10-m mean wind speed for the entire period (36 yr) have been used. Regarding the future projections, we use the regional simulations database available from the ENSEMBLEbased predictions of climate changes and their impacts (ENSEMBLES) European project (Table 1; Hewitt 2004; further information at These models were nested within four different driving general circulation models and were run from 1951 to 2100 under the SRES A1B scenario. The experiments were performed using a 25-km horizontal grid-length resolution that spans Europe and includes the eastern part of the Atlantic, northern Africa, and western Asia. Daily-averaged simulated variables for each model have been bilinearly interpolated to LEPA from the four nearest grid points (Akima 1978, 1996). Even if dynamical downscaling improves the representation of regional features in climate projections, some important local inaccuracies still remain owing to insufficient resolution and the uncertainties in the representation of small-scale forcings and processes (e.g., clouds, convection, boundary layer, radiative transfer). Several procedures exist to adjust RCM projections, taking local forcings into consideration. Two straightforward corrections consist of (i) adding the climatological difference between future and control climate scenario simulations to an observed baseline (the so-called delta method) and (ii) removing the bias from future simulation by applying the climatological difference between the observed and control data (the unbiasing method; Déqué 2007). These techniques assume that the variability in the climate scenario remains unchanged in the first case and that the RCM variability is perfect in the 1FEBRUARY 2012 A M E N GUAL ET AL. 943 second case, two important assumptions. The application of the quantile quantile mapping transformation is more flexible than the previous methods and is a procedure that has been widely used for correcting biases in the simulated meteorological variables (Wood et al. 2004; Reichle and Koster 2004; Déqué 2007; Boé et al. 2007). Within this context, we present a new quantile quantile calibration method based on a nonparametric function that amends mean, variability, and shape errors in the simulated cumulative distribution functions (CDFs) of the climatic variables. The procedure consists of calculating the changes, quantile by quantile, in the CDFs of daily RCM outputs between a 15-yr control period and successive 15-yr future time slices. These changes are rescaled on the basis of the observed CDF for the same control period, and then added, quantile by quantile, to these observations to obtain new calibrated future CDFs that convey the climate change signal. We have chosen periods of 15 yr owing, first of all, to the temporal limitation of the observed database (36 yr): we have split the daily series into an early control period for the calibration task ( ; baseline) and a later interval for validation purposes ( ); the remaining years (i.e., ) are used to perform a robustness test of the method (section 3b). Second, we consider a length of 15 yr to be a compromise between series large enough to have climatological meaning the statistical sample is N and short enough to permit, by comparing the simulated CDFs of successive 15-yr intervals starting in 2010, an effective isolation of any climate change signal along the twenty-first century. Recalling that our control period extends from 1973 to 1987 and that the future periods comprise all subsequent 15-yr intervals after 2010, the statistical adjustment (developed in detail in the appendix) can be written as the following relationship between the ith ranked value p i (projected or future calibrated), o i (control observed or baseline), s ci (raw control simulated), and s fi (raw future simulated) of the corresponding CDFs (Fig. 2): where p i 5 o i 1 gd 1f D9 i, (1) D i 5 s fi 2 s ci, (2) D5 å N D i i51 N N å(s fi 2 s ci ) 5 i51 N 5 S f 2 S c, and (3) D9 i 5D i 2 D, (4) FIG. 2. Graphical sketch of the Q Q adjustment. The CDFs of the mean temperatures are shown for the observed control (OBS ), raw control (RCM ), and future (RCM ) simulated, and calibrated or projected (PRJ ) data. The statistical correction is illustrated between the 15-yr past ( ) and future ( ) periods. Vertical lines denote mean values for raw control (S c ) and future (S f ) simulated periods. and g 5 å N i51 å N i51 o i s ci 1, A N 1, 5 O and (5) S c A N f 5 s O 5 IQRj O. (6) s Sc IQRj Sc As surrogates of the population variability, IQRj O and IQRj Sc in (6) are the interquartile ranges of the observed and raw control
Search
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x