Temperature and Precipitation Variability and Trends in Northern Spain
269
low correlation between NAO and winter temperature for Spain has been reported
by other researchers (Pozo et aI., 1999). As shown above spring temperature is
highly correlated (positive) with Gibraltar pressure, and can also be explained in
terms of the high solar heating under clear conditions during the anti-cyclone
patterns and a cooling during cyclonic activity, but the radiative mechanism is not
enough to provoke a low minima during the clear nights. This is also applicable to
autumn and annual data. Meanwhile low negative correlations in summer would be
unexpected. This maybe due to the stable general high-pressure conditions and the
weakening circulation. As some authors argue (Wallace et aI., 1995, 1996), the
temperature variability during summer regions would be caused by radiati ve forcing
instead of circulation because this factor is very weak during the warm season. But
as Figure 4 shows, these relationships have changed markedly over time, reflected
in an apparent increase in the influence of pressure.
The Azores high appears to have been more intense and to have been desplaced
towards the east and the north since 1980 (Beniston et aI., 1994), and it could be
responsible for the increases in temperatures during these years. In any case, such
pressure or Atlantic circulation conditions are not enough to explain the temperature
variability on scales of decades or years. For example, there is a good agreement
with the pressure values over the Mediterranean sea at 40oN; high-pressure values
correspond to cooler periods, and low-pressure values to warmer periods
(Makrogiannis and Sashamanogluo, 1990).
A partial explanation of the increase found is the so-called urban island effect.In
the present work the most part of the localities correspond to urban sites, from large
cities such as Madrid and Barcelona, to rural localities such as Monflorite and
Tablada. Most are small cities (41.5% of these localities have less than 100.000
inhabitants, and 22% less than 50.000 inhabitants). It is true that urban heating is
also present in small localities as demonstrated by Karl et al.(1988), but this
conclusion loses validity for cities with a very long history, as it is in this case.
Urban growth has not had a constant rate, and thus, it is not possible to consider a
linear trend. At least in Spain, growth has been greater during this century,
especially since 1960, when the cities began to develop modern features, such as the
presence of asphalt and industry. In addition, the number of cities with more than
100.000 inhabitants rose in the 1970s, although the warming began before
1920.Also, the cooling in the 1970s does not seem masked by the urban heating.
According to Jones et al. (1989), at least at hemispheric scale, the effect of the
urbanism is less than 0.1 0c. Given the general agreement between their series and
ours, the urban island effect is perhaps responsible for the temperature increase in
Spain at a similar rate.
The magnitude of the urban heating effect depends of several variables: size and
behaviour of the urban nucleus, topography, local industry and the meteorological
conditions (Landsberg, 1981). Kozuschoswki et aI., (1994) estimate the magnitude
of this heating for the industrial city of Krakow in 0.5 °C during the last 40 years.
Beniston et al. (1994) consider that the urban island effect is very small for cities
like Zurich and Lugano, taking into account the global increase of the temperature
269
low correlation between NAO and winter temperature for Spain has been reported
by other researchers (Pozo et aI., 1999). As shown above spring temperature is
highly correlated (positive) with Gibraltar pressure, and can also be explained in
terms of the high solar heating under clear conditions during the anti-cyclone
patterns and a cooling during cyclonic activity, but the radiative mechanism is not
enough to provoke a low minima during the clear nights. This is also applicable to
autumn and annual data. Meanwhile low negative correlations in summer would be
unexpected. This maybe due to the stable general high-pressure conditions and the
weakening circulation. As some authors argue (Wallace et aI., 1995, 1996), the
temperature variability during summer regions would be caused by radiati ve forcing
instead of circulation because this factor is very weak during the warm season. But
as Figure 4 shows, these relationships have changed markedly over time, reflected
in an apparent increase in the influence of pressure.
The Azores high appears to have been more intense and to have been desplaced
towards the east and the north since 1980 (Beniston et aI., 1994), and it could be
responsible for the increases in temperatures during these years. In any case, such
pressure or Atlantic circulation conditions are not enough to explain the temperature
variability on scales of decades or years. For example, there is a good agreement
with the pressure values over the Mediterranean sea at 40oN; high-pressure values
correspond to cooler periods, and low-pressure values to warmer periods
(Makrogiannis and Sashamanogluo, 1990).
A partial explanation of the increase found is the so-called urban island effect.In
the present work the most part of the localities correspond to urban sites, from large
cities such as Madrid and Barcelona, to rural localities such as Monflorite and
Tablada. Most are small cities (41.5% of these localities have less than 100.000
inhabitants, and 22% less than 50.000 inhabitants). It is true that urban heating is
also present in small localities as demonstrated by Karl et al.(1988), but this
conclusion loses validity for cities with a very long history, as it is in this case.
Urban growth has not had a constant rate, and thus, it is not possible to consider a
linear trend. At least in Spain, growth has been greater during this century,
especially since 1960, when the cities began to develop modern features, such as the
presence of asphalt and industry. In addition, the number of cities with more than
100.000 inhabitants rose in the 1970s, although the warming began before
1920.Also, the cooling in the 1970s does not seem masked by the urban heating.
According to Jones et al. (1989), at least at hemispheric scale, the effect of the
urbanism is less than 0.1 0c. Given the general agreement between their series and
ours, the urban island effect is perhaps responsible for the temperature increase in
Spain at a similar rate.
The magnitude of the urban heating effect depends of several variables: size and
behaviour of the urban nucleus, topography, local industry and the meteorological
conditions (Landsberg, 1981). Kozuschoswki et aI., (1994) estimate the magnitude
of this heating for the industrial city of Krakow in 0.5 °C during the last 40 years.
Beniston et al. (1994) consider that the urban island effect is very small for cities
like Zurich and Lugano, taking into account the global increase of the temperature