Riassunto del Workshop scientifico "Climate Change at High Elevation Sites: Emerging Impacts HIGHEST II"

The rationale: The conference built on the momentum created by the first conference on Climate Change at High Elevation Sites held in Wengen, Switzerland in 1995. As a follow-on activity, 40 scientists came together to discuss new findings since HIGHEST I, i.e. 1) to review recent climatic trends at different high elevation sites around the world, 2) to evaluate the utility and fidelity of ecological indicators of climate change at high elevations, and 3) to evaluate the impact of climate change at high elevation sites on Earth surface processes, especially biotic and abiotic (e.g., water) resources.

The abstract volume is available at http://www.nccr-climate.unibe.ch

Here a few selected highlights (not exclusive and not complete):

"Global moistening" rather than "Global warming"

Several speakers (e.g. Schaer et al.) made a clear statement that temperature and precipitation must always be regarded as coupled variables. Temperature changes (e.g. 'Global Warming') always depend on the fraction of energy being converted into latent or sensible heat flux. This must be considered when rates of temperature changes in different locations and altitude levels are assesses, when changes in lapse rates are compared, freezing height levels of the free atmosphere and ground temperature are computed etc. Schaer et al showed that the atmospheric water vapor content increases with global warming after Clausius-Clapeyron (~ 6%/K) whereas global precipitation increase is only half (1-3%/K) of Clausius-Clapeyron. Thus the water content of the atmosphere increases substantially with global warming, and 'Global Moistening' is the more appropriate term than 'Global Warming'.

Alpine climate and NAO

Observed trends during the 20th century show that 1) mean winter precipitation has significantly increased (+30%) particularly in the NW Alps, 2) intense precipitation events increased, and 3) mean and peak runoff and river discharge changed while the summer conditions remained remarkably stable. Since the observed trend in winter precipitation is about three times stronger than what has been used a decade ago for the climate change scenarios for the 21st century (+10% precipitation) it is suggested that also the impact on the Alpine hydrological cycle will be much greater that what is expected for the next 100 years based on model results from experiments and scenarios 10 years ago. (e.g., NFP 31, Schaer et al.). Future climate scenarios suggest with remarkable agreement that the observed trends for the 20th century (mild and moist winters, increased floods) will continue, but models also suggest that decreasing summer precipitation in the Southern Alps and the Mediterranean will increase the likelihood for droughts. There are only few observations for that effect.

Whereas European temperature and precipitation fields are highly correlated with the NAO index, the correlation between Alpine temperature and precipitation with NAO index are high only for distinct periods in winter. However, the northern and southern Alps undergo different NAO influences (Wanner et al.). The mechanism describing the dynamical influence of NAO on Alpine climate is seen in the varying position of the active polar front jet. In the zonal (positive NAO) mode, the Alps are located SE of the exit zone of the jet, which leads to a remarkably reinforced cross-isobar NW-SE mass transport, and ultimately in anticyclonic and warm winter weather over the Alpine cone.

Temperature trends and freezing height levels

Rates of temperature changes in the Alps (Beniston) and Himalayas (e.g. Shresta) exceed global average values. Vuille reported new findings from the Andes. Mean annual temperatures (surface observations) in the northern tropical Andes of Colombia increased by as much as 1.6°C in 25 years (1966 to 1990). Further to the south in Peru, Bolivia, Ecuador, the overall rate of temperature increase is not as high but the rate has more than tripled during the last 25 years (1975 - 1990) compared to the decades before (1939-1975). In Chile and parts of Argentina the warming rates during the last three decades (1960-1992) has doubled compared to the period between 1933 to 1960 (Vuille et al.). Little is known about precipitation changes. Vuille et al. found no significant P decrease over the last tree decades. However, an increase in March-April-May precipitation goes along with an increase in convective cloudiness in the eastern Andes and westernmost Amazonia while cloudiness in subtropical South America decreased. A more active Hadley circulation may offer a possible explanation for this observation.

In the free atmosphere (the mid-troposphere is about the altitude of glaciers), the temperature trend is little or not as clear (Seidel, Diaz, Vuille). Surface observations and free atmosphere soundings report different signs and changes for lapse rates in different parts of the world (tropics and mid- latitudes). These differences and the relationship between surface and free-air temperatures in different regions are poorly understood. A comprehensive description of the problem comparing free atmosphere and surface temperatures is given by Seidel and Free (Abstract Volume, 41). Research is needed to better understand the apparent discrepancy between massive glacier retreat, increased surface temperatures at high elevation sites but the lack of large-scale warming of the lower- and mid-troposphere. Particular attention should be paid to luv-lee effects, seasonality, tropical SSTs and ENSO, and the influence of large mountain massifs and plateaus compared to isolated peaks.

Glaciers and snow:

Recent surveys from Kilimanjaro show strongly negative mass balances during the last years (Kaser, Thompson et al.).

About 100 glaciers disappeared in the Alps since AD 1850. Observed glaciers retreats in the mid-latitudes are between 26- 35% in area and up to 50% in volume (Meier, Föhn). Recent retreats of South Cascade Glacier have been well outside the range of all fluctuations of this glacier during the past 5000 years.

Again, glacier retreat must be regarded as a function of temperature and precipitation (among other very important factors such as e.g. seasonality) and it has been shown that the modes of mass balance variability (summer and winter) are well related to variations in mesoscale atmospheric circulation (Meier). Interestingly, most recent observations revealed massive advances of glaciers in the Cordillera Blanca of Peru since 1998 (Kaser), which might be related to the observed increase in cloudiness and convection over the eastern Amazon basin (Vuille).

Special emphasis was also given to tropical ice cores and the interpretation of chemical species and isotopic compositions. Results from three cores from Tibet (Dunde, Gulya and Dasuopu, Thompson et al.), five Andean ice cores (Huascaran, Quelccaya, Sajama, Illimani, Chimborazo, Tapado (Thompson et al., Schotterer et al.) and Africa (Kilimanjaro; Thompson et al.) were compared. A particular problem is the question if seasonality is preserved in the ice cores, if the record is complete (e.g. high sublimation rates may erase depositions of a year) and how to detect possible hiatuses in the cores. A new method of reconstructing original mass balance terms was proposed (Schotterer, Ginot et al.). This is mandatory, if measured net accumulation rates in the ice core are interpreted in terms of precipitation (total accumulation) and ultimately climate. All of the tropical ice cores show significant isotopic enrichment in the 20th century.

Hydrology

The most robust projection for changes in river discharge is that global warming would result in earlier and smaller snowmelt-fed peak runoffs. Discharge data around the world show this effect in mountain areas of the high- and mid-latitude northern Hemisphere and (less robust) on the southern hemisphere (M. Dettinger). Peak discharge is earlier, and AMJ discharge (a measure for late- snow discharge) is a significantly smaller fraction of the annual total (Eastern Europe, Russia, Canada, Rocky Mountains and Mississippi River). This means that the natural water storage (snow, ice) capacity decreases dramatically. The changes are most pronounced in mid-altitude mountain ranges, i.e. middle mountains are more sensitive to such changes compared with high-elevation areas.

Vegetation and biodiversity

20th century increases in tree line were reported for the Urals (Moiseev): In the South Ural 60 -80 m (observation period 1929 to 2000) and in the North and Polar Urals 40 m (observation period 1957 - 2000). This is attributed to warming (observed on average 1.4°C over the last century) and moistening. Cold season warming (+3°C in the South Ural, +4.3°C in the North and Polar Urals compared to the mid-19th century) is stronger than the warm season warming. Also precipitation rates increased substantially. As a result of the upward expansion of forests, tundra areas in the South Urals decreased by 10-30% over the last century (preliminary estimates), and are thought to decrease by 40-70% if T rises by 3°C.

Christian Koerner argued convincingly that surface air temperatures and soil temperatures in the root zone (and not CO2!) are the most important factors controlling the upper tree line elevation.

Tree ring chronologies from seven mountain ranges in Arizona and New Mexico (Swetnam et al.) show a rise in growth since 1976 that is unprecedented over the period of the last 900 years. This is thought to be due to exceptionally warmer and wetter conditions. Also other ecological processes (forest fire and tree recruitment) changed since 1976.

New dendroclimatological data (Villalba, et al.) from the Southern Andes also show how unusual the 20th century climate was in the context of the last 400 years. Southern Patagonian temperatures (from tree rings) since the mid-1970s were higher than what was observed in tree rings over the last 350 years.

Also treeing studies from Mongolia (D'Arrigo, Jacoby et al.) suggest that warmest temperatures over the past thousand years are found in the 20th century. The 1999 tree ring width has probably the highest value over the past millennium.

Also methodological problems were highlighted. Key-issues concern a) the differences in the ring parameters being used, b) differences in the climate parameter modelled, c) inherent species-related differences, d) actual climate differences at the site, e) differences in the statistical procedures used for the reconstruction (Hughes, Luckman, Swetnam, Villalba, Jacoby et al).

25% of the European vascular plants are restricted or predominantly occurring in the alpine zone (above tree line), although this area covers only 3% of the area. Comparison of recent vascular flora with surveys in the late 19th and early 20th century on more than 300 alpine summits show that the observed warming during the 20th century resulted in an increase in species richness in the nival zone (Grabherr et al.). GLORIA- Europe Global Observation Research Initiative in Alpine Environments started on January 1st 2001. GLORIA-Europe is the pilot project for a global network. Camille Parmesan presented examples of butterfly diversity at high elevation sites in western USA and Europe. Some populations showed massive increases, others got extinct. A main response to climate change is also the redistribution of species and communities. For example the mean location of Edith's Checkerspot (Western USA) populations shifted 124 m upward during the 20th century, which matches well the observed mean annual T trend. The upper level of the Edith's Checkerspot population seems to be stable. Other mountain species (e.g. Apollo's butterfly) got extinct on all tops of the Jura Mountains (Europe) lower than 850 m in the early 1970s whereas the populations remained present on mountains higher that 850 m.