condensates, the different precipitation episodes can be categorized according to their intensity, duration, frequency,
and by type (stratiform, like those caused by depressions in
the mid-latitudes, or convective, like heavy rains and tropical
cyclones in the intertropical convergence zone). These
characteristics depend to a large extent on local temperature
and weather conditions (wind speed and direction, pressure,
humidity, evaporation). It is thus clear that a change in any
of these parameters will affect the hydrological cycle as a
whole.
Throughout the twentieth century, for example, annual
precipitation increased on the eastern side of the South and
North American continents. In contrast, a significant rainfall
deficit has been observed in south and west Africa, as well as
in the Sahel. In northwestern India, an increase of about 20%
was observed for the period 1901–2005, despite a sharp
decrease between 1979 and 2005. However, changes in
precipitation are hard to measure using the existing records,
and there is only medium confidence that observed trends
are due to the anthropogenic influence.
Observed trends in relative humidity (i.e. air
humidity/saturation humidity) suggest that these have
remained constant through the tropospheric column down to
the surface. However, if the amount of water vapor at saturation increases and the relative humidity remains constant,
then this means that the absolute humidity (and thus the
amount of water vapor) has increased in the atmosphere.
Observations indicate that tropospheric water vapor has
increased by about 3.5% over the past 40 years, which is
consistent with the observed temperature change of 0.5 °C
over the same time period. Climate models confirm these
empirical observations: a warmer climate leads to increased
moisture content in the atmosphere and more intense precipitation events (although total precipitation over a full year
is reduced) and therefore flooding is more likely to occur.
Thus, in winter, it is observed that for most of the extratropical land surface areas of the northern hemisphere, the
greatest precipitation is linked to higher temperatures. Conversely, in areas with low rainfall, such as the Mediterranean
basin, rising temperatures are associated with a higher risk of
drought. This general intensification of the hydrological cycle
can be summarized by: wet gets wetter, dry gets drier.
Added to these complex phenomena is the variability in
atmospheric circulation. In Sect. “Climate Variability”, the
modes of atmospheric variability will be examined in greater
detail. However, we already know that fluctuations in
atmospheric circulation over the North Atlantic brought
heavy rainfall in the 1990s to northern Europe, and, in
contrast, led to a drying-up of the Mediterranean basin. In
addition, the severe Sahelian drought over more than
20 years (1970–1990) was linked to changes in both atmospheric circulation and surface ocean temperatures in all
three of the Pacific, Indian and Atlantic basins. Although this
trend towards drought still persists, it has become less pronounced since the early 1990s.
There are numerous uncertainties surrounding the determination of the hydrological cycle variables. This is due to a
lack of data for some regions (for example, Canada,
Greenland and Antarctica, some desert regions such as the
Sahara, the Tibetan plateau and over the oceans), and to the
fact that accurate measurements have only been available for
a very short time. In addition, it is very difficult to measure
precipitation rates and to accurately quantify their changes at
the global and regional scales.
In situ measurements are affected by atmospheric conditions (e.g. the effect of strong winds especially on snowfall).
Spatial observations, on the other hand, provide only
instantaneous measurements and are affected by the uncertainties associated with the algorithms used to convert
radiometric measurements into precipitation rates. Because
of these difficulties, climatologists explore the coherence
between the whole set of complementary variables associated with the hydrological cycle. One way to represent
precipitation changes over the past century or over decades
is to calculate the Palmer Drought Severity Index. This index
is a measure of drought, in other words, the accumulated
surface soil moisture deficit compared to average local
conditions. It is based on recent rainfall and atmospheric
humidity (determined from temperatures). In general, analyzes of the last century suggest a trend towards drying for
much of Africa, southern Eurasia, Canada and Alaska. Other
direct or indirect measurements (e.g. from river flow estimations or ocean salinity measurements) show that during
the twentieth century precipitation has generally increased
on land surfaces between 30°N and 85°N, but significant
reductions have been observed over the last thirty or forty
years between 10°S and 30°N.
At present, the main challenge is to determine the interannual variations and trends in precipitation changes over
the oceans. Global averages are often unrepresentative and
mask large regional disparities. However, particularly pronounced droughts in the last 30 years, as well as heavy
rainfall events in many regions clearly illustrate an intensification of the hydrological cycle.
Extreme Weather Events
Extreme weather events result from exceptional fluctuations
of a climate variable, and are generally associated with
significant impacts on society and the environment. As
illustrated by the frequent exposure by the media of the most
impressive events, the statistical evolution of climate
extreme characteristics (frequency, amplitude) is of major
concern for current climate change, especially in the development of adaptation strategies. The term ‘weather extreme’
448
S. Charbit et al.
and by type (stratiform, like those caused by depressions in
the mid-latitudes, or convective, like heavy rains and tropical
cyclones in the intertropical convergence zone). These
characteristics depend to a large extent on local temperature
and weather conditions (wind speed and direction, pressure,
humidity, evaporation). It is thus clear that a change in any
of these parameters will affect the hydrological cycle as a
whole.
Throughout the twentieth century, for example, annual
precipitation increased on the eastern side of the South and
North American continents. In contrast, a significant rainfall
deficit has been observed in south and west Africa, as well as
in the Sahel. In northwestern India, an increase of about 20%
was observed for the period 1901–2005, despite a sharp
decrease between 1979 and 2005. However, changes in
precipitation are hard to measure using the existing records,
and there is only medium confidence that observed trends
are due to the anthropogenic influence.
Observed trends in relative humidity (i.e. air
humidity/saturation humidity) suggest that these have
remained constant through the tropospheric column down to
the surface. However, if the amount of water vapor at saturation increases and the relative humidity remains constant,
then this means that the absolute humidity (and thus the
amount of water vapor) has increased in the atmosphere.
Observations indicate that tropospheric water vapor has
increased by about 3.5% over the past 40 years, which is
consistent with the observed temperature change of 0.5 °C
over the same time period. Climate models confirm these
empirical observations: a warmer climate leads to increased
moisture content in the atmosphere and more intense precipitation events (although total precipitation over a full year
is reduced) and therefore flooding is more likely to occur.
Thus, in winter, it is observed that for most of the extratropical land surface areas of the northern hemisphere, the
greatest precipitation is linked to higher temperatures. Conversely, in areas with low rainfall, such as the Mediterranean
basin, rising temperatures are associated with a higher risk of
drought. This general intensification of the hydrological cycle
can be summarized by: wet gets wetter, dry gets drier.
Added to these complex phenomena is the variability in
atmospheric circulation. In Sect. “Climate Variability”, the
modes of atmospheric variability will be examined in greater
detail. However, we already know that fluctuations in
atmospheric circulation over the North Atlantic brought
heavy rainfall in the 1990s to northern Europe, and, in
contrast, led to a drying-up of the Mediterranean basin. In
addition, the severe Sahelian drought over more than
20 years (1970–1990) was linked to changes in both atmospheric circulation and surface ocean temperatures in all
three of the Pacific, Indian and Atlantic basins. Although this
trend towards drought still persists, it has become less pronounced since the early 1990s.
There are numerous uncertainties surrounding the determination of the hydrological cycle variables. This is due to a
lack of data for some regions (for example, Canada,
Greenland and Antarctica, some desert regions such as the
Sahara, the Tibetan plateau and over the oceans), and to the
fact that accurate measurements have only been available for
a very short time. In addition, it is very difficult to measure
precipitation rates and to accurately quantify their changes at
the global and regional scales.
In situ measurements are affected by atmospheric conditions (e.g. the effect of strong winds especially on snowfall).
Spatial observations, on the other hand, provide only
instantaneous measurements and are affected by the uncertainties associated with the algorithms used to convert
radiometric measurements into precipitation rates. Because
of these difficulties, climatologists explore the coherence
between the whole set of complementary variables associated with the hydrological cycle. One way to represent
precipitation changes over the past century or over decades
is to calculate the Palmer Drought Severity Index. This index
is a measure of drought, in other words, the accumulated
surface soil moisture deficit compared to average local
conditions. It is based on recent rainfall and atmospheric
humidity (determined from temperatures). In general, analyzes of the last century suggest a trend towards drying for
much of Africa, southern Eurasia, Canada and Alaska. Other
direct or indirect measurements (e.g. from river flow estimations or ocean salinity measurements) show that during
the twentieth century precipitation has generally increased
on land surfaces between 30°N and 85°N, but significant
reductions have been observed over the last thirty or forty
years between 10°S and 30°N.
At present, the main challenge is to determine the interannual variations and trends in precipitation changes over
the oceans. Global averages are often unrepresentative and
mask large regional disparities. However, particularly pronounced droughts in the last 30 years, as well as heavy
rainfall events in many regions clearly illustrate an intensification of the hydrological cycle.
Extreme Weather Events
Extreme weather events result from exceptional fluctuations
of a climate variable, and are generally associated with
significant impacts on society and the environment. As
illustrated by the frequent exposure by the media of the most
impressive events, the statistical evolution of climate
extreme characteristics (frequency, amplitude) is of major
concern for current climate change, especially in the development of adaptation strategies. The term ‘weather extreme’
448
S. Charbit et al.