Réf. ALP-IMP 2006 - R

Référence bibliographique complète
Multi-centennial climate variability in the Alps based on Instrumental data, Model simulations and Proxy data. ALP-IMP, 2006, Final report for RTD-project. 102 p.

Mots-clés
Climate variability, instrumental data, reconstruction methods, modeling, Greater Alpine Region.

Organismes / Contact
Partenaires
Project coordination: Reinhard Böhm (Central Institute for Meteorology and Geodynamics, Vienna, Austria).
http://www.zamg.ac.at/ALP-IMP
University of East Anglia, GKSS Forschungszentrum, Universität Heidelberg, Consiglio Nazionale delle Richerche, World Glacier Monitoring Service/University Zurich, Laboratoire des Sciences du Climat et de l'Environment, Swiss Federal Research Institute, Wood Biology Research Teem/University of Agricultural Sciences, University of Innsbruck.

Principaux rapports scientifiques sur lesquels s'est appuyé le rapport
 

(1) - Paramètre(s) atmosphérique(s) modifié(s)
(2) - Elément(s) du milieu impacté(s)
(3) - Type(s) d'aléa impacté(s)
(3) - Sous-type(s) d'aléa
Temperature, Precipitation, Sunshine, Humidity, Circulation patterns Glaciers    

Pays / Zone
Massif / Secteur
Site(s) d'étude
Exposition
Altitude
Période(s) d'observation
Europe Alps       Last 1000 years

(1) - Modifications des paramètres atmosphériques
Reconstitutions
Tree-ring records
The MXD-RCS preserves high to low frequency information and explains 60% of Alpine temperature variations back to 1818, however, with a clear weighting towards high frequency variation. Both proxies reveal warm conditions from before AD 1000 into the 13th century, followed by a prolonged cool period, reaching minimum values in the 1820s, and a warming trend into the 20th century. The high temperatures in the 10th and 13th century, comparable to those of the last decade, confirm the putative Medieval Warm Period. The cooling from ~1300-1820, relative to the 20th century, reflects to the so-called Little Ice Age. With 2003 being the warmest summer over the past 1250 years (but closely followed by the pre-industrial, medieval summer AD 970), the proxies capture the full range of the instrumental measurements. The MXD chronology provides an annual (decadal) temperature amplitude of 6.4 (3.1)°C. The new GAR reconstructions suggest that summer temperatures during the last decade are unprecedented over the past millennium.

Investigations on precipitation changes, however, are much more challenging because of the more spatially heterogeneous nature of precipitation itself and the less uniform signal in low elevation tree-ring samples.

1000 years of climate in the GAR
The early instrumental temperature series for the GAR showed a marked warm period during the 1790s and 1800s (comparable to the mean warmth of the 1980s, but below that for the years 1990-2005). Following the 1800s, temperatures cooled markedly in the 1810s, were warm again in the 1820s and then cooled to their lowest levels from the 1830s right through to the early-20th century. Over the period 1760 to 2003, the tree-ring density data indicate that the warmest summer occurred in 2003 (in agreement with the instrumental data).

The coolest summer as indicated by the trees was in 1816. This is not consistent with the instrumental record, where the summer of 1816 is only one among the coldest three summers that occurred during the 1760-2003 period.

The long tree-ring density reconstruction since 755 shows warmer summers in the Medieval period, with cooler summers between about 1350 and 1820 and the warmest of all in the last 20 years. Particularly warm decades were recorded by the trees during the 960s to 980s, the 1200s to 1220s and the recent 25 years. Of particular note are the cooler summers during the 1040s to 1060s. Cool temperatures were also recorded, particularly during the period from 1400 to 1710 and during the 1810s.

Warm summers were evident around 1500 and 1800. These are partly confirmed by the ring-width-based reconstruction, but the two reconstructions are markedly at odds with each other during the first two centuries of the millennium, particularly with respect to the trend over that period.
Observations
200 years of climate in the GAR
The annual mean temperature evolution in the GAR in the instrumental period can be characterized by two main sections. A 100-years period from 1790 to 1890 with a cooling of -0.97K was followed by a 116 years warming by +1.48K. Compared to the global trend from 1890 to 2005 of +0.74K, the GAR has warmed nearly twice as much. Most of the stronger GAR-warming has been caused by two outstanding periods mainly: in the 1890s and in the recent 20 years. From 1900 to the late 1980s the GAR behaved rather similar to the global evolution. A warmth of two decades (from 1790 to 1810), followed by a sharp cooling in the 1810s has been highlighted.

The MSLP-patterns play a significant role in the GAR in winter, much less in summer. High elevation temperatures are markedly linked with the Northern Hemisphere zonal circulation, whereas the low elevation temperature field is associated more with the circulation over the NE-Atlantic. North of the Alps, a British Isles-centred pressure pattern plays the principal influence on winter precipitation, whereas the Mediterranean subregions are dominated by NAO. The impact of the ENSO phenomenon on GAR climate is weak.
Modélisations
Climate variability in the GAR
In regard to low frequency variability, decadal scale evolutions resulted to be rather similar for the entire GAR for temperature. Especially the non existing difference between high and low elevations is of special interest as well as the identical trends between rural and urban sites.

Precipitation shows the most outstanding regional trend differences of all climate elements at decadal to centennial scale. Particularly between the NW versus the SE subgroup – obviously caused by the obstacle of alpine chain – the trends of the last 150 years even show opposite sign with a 10% increase in the NW and a 10% decrease in the SE of the GAR. The most antagonistic pair of seasonal precipitation (NW-winter vs. SE-autumn) show very long-term stable wetting/drying trends over 120/180 years, but also both long-term trends have abruptly changed into their opposite in recent times. Winter precipitation is decreasing again since 1980, autumn precipitation increasing in the SE since 1990.

The most interesting findings are the different 20th century evolutions of high versus low elevation annual sunshine totals. There was a clear trend of significant “brightening” at high elevations (2000-3500m altitude) in both sections of the 20th century whereas the low elevations (below 1000m asl.) show weaker to not significant sunshine trends. Cloudiness trends confirm the decadal scale sunshine features, at centennial scale there are some yet not well understood misfits or perhaps some remaining homogeneity problems.

Vapour pressure, a measure for the absolute humidity content of the air, follows closely the general increase of air temperature and also some of the decadal scale peculiarities at low as well as for high altitudes. Relative humidity, on the other hand has reacted differently on the warming of the past 120 years. At low elevations a long-term 7%-drying from 1880 to 2005 has been highlighted. At the elevation of the Alpine summit observatories, on the other hand, the long-term drying trend has been considerably less marked due to their closer coupling to the maritime source regions.
Hypothèses
 

Informations complémentaires (données utilisées, méthode, scénarios, etc.)
Tree-ring records
The project led to the development of a spatially dense Greater Alpine Region (GAR) tree ring network including more than 400 ring width (TRW) and more than 130 density chronologies from 6 main species (Abies alba, Larix deciduas, Picea abies, Pinus cembra, Pinus nigra and Pinus sylvestris). The following sub-regions of the GAR network have been defined for different aspects of climate reconstruction: Valais and Engadin in the Central Swiss Alps, Tyrolean Central Alps, Dachstein, Northern Limestone Alps, Vienna basin (all Austria).

A newly developed method known as “Regional Curve Standardisation” (RCS) has been used to produce chronologies. For calibration/verification statistics, various regression models are applied including different periods, seasonalities, and wavelengths. To understand climatic extremes and their temporal distribution, a larger Alpine network of high elevation temperature sensitive tree sites was analyzed, preserving the relative frequency and magnitude of extreme events.

1000 years of climate in the GAR
Tree ring reconstructions are annually resolved and dating is assured through comprehensive crossdating of the developed chronologies. Ice-core-based reconstructions using isotope measurements also have potential annual resolution, but dating is more problematic and cannot be considered exact. Emphasis here is, therefore, initially placed on the tree-ring results, with confirmation of longer-timescale changes sought through graphical comparisons. Finally, a number of earlier reconstructions, principally from documentary-based reconstructions since about 1200, are included to help understand the reconstructions for the GAR.

200 years of climate in the GAR
Instrumental climate records collected from 242 sites have been used: an early period from approximately 1760 to 1850 with a limited number of series, a fully developed network in the 20th century and a transition period in the second part of the 19th century. The instrumental climate database HISTALP, which existed before the start of the project, was completely re-analyzed. On the one hand the additional data allowed a better testing and adjusting of already pre-homogenised series. On the other hand some more sophisticated techniques had emerged. Hundreds of inhomogeneities and thousands of outliers were detected and were subsequently eliminated. The elimination of inhomogeneities made the dataset fit for analysis of climate trends, the correction of outliers now allows for a correct analysis of climate extremes. Before about 1820, the instrumental data are warmer than the tree-ring density record would imply. The difference is of the order of 0.5°C, suggestive of residual homogeneity problems in the early instrumental records.

The influence of large-scale atmospheric circulation on GAR temperature and precipitation was studied using some continental to global scale leading MSLP-patterns derived from the EMULATE monthly gridded sea level pressure dataset (EMSLP, 1850-2003), the monthly North Atlantic oscillation index (NAO, 1821-2004), the Arctic Oscillation index (AO, 1899-2002), the Southern Oscillation Index (SOI, ENSO, 1850-2004 and a tree-ring based reconstruction 1706-1977) and the Nino3 Index (1408-1978).

Climate variability in the GAR
A high-resolution regional simulation has been performed with the regional model REMO (Jacob and Podzun 1997) over the whole of Europe for the period 1958 to 1998. The simulation with a resolution of 1/6 deg (about 17 km) on 20 vertical levels is driven by the 1.125 deg resolution ERA40 reanalysis.

(2) - Effets du changement climatique sur le milieu naturel
Reconstitutions
Observations
 
Modélisations
It has been shown that current rates of glacier wastage by far exceed the historical changes and that deglaciation of entire mountain ranges within the coming decades must be taken into account.
Hypothèses
 

Sensibilité du milieu à des paramètres climatiques
Informations complémentaires (données utilisées, méthode, scénarios, etc.)
 
A suite of models with varying complexity was developed. A strong focus was on the development of simple but robust models, that can be widely applied and make efficient use of the climatic time series compiled within ALP-IMP.

(3) - Effets du changement climatique sur l'aléa
Reconstitutions
 
Observations
 
Modélisations
 
Hypothèses
 

Paramètre de l'aléa
Sensibilité du paramètres de l'aléa à des paramètres climatiques
Informations complémentaires (données utilisées, méthode, scénarios, etc.)
 
 

(4) - Remarques générales
 

(5) - Préconisations et recomandations
Destinataires et portée du rapport Scientific community
Types de recommandations et / ou préconisations