Geoscience Reference
In-Depth Information
consists of ash (
Fraxinus angustifolia
), willows (
Salix
atrocinerea
and
S. salviifolia
), poplars (
Populus nigra
)
and introduced African tamarisk (
Tamarix africana
).
Field measurements were made at a 1
×
1
km study site in the middle reaches, selected for
its location and ease of access (Figure 22.1a).
The channel in the study site had a bedslope of
0.32%, a mean width of approximately 20 m and a
maximum depth of 1.4 m. The substrate comprised
mainly sand, gravel and pebbles. Maximum
altitude in the study site was 243 m. Changes in
land use and stream morphology at the study site
were assessed using aerial photographs taken in
1958 and 2007 (Figure 22.1b).
the headwaters, middle and lower reaches. Data
used for the SWAT model were taken from
a number of sources, including: (i) 1:100 000
scale soil map (Gon¸alves
et al.
, 2005); (ii)
Corine Land Cover 2000 at 1:100 000 scale
(http://www.igeo.pt/); and (iii) the Alentejo Digital
Terrain Model, provided by the Administra¸ ao
da Regi ao Hidrogr afica do Alentejo I.P. Field
measurements included data from a weather
station specially installed in the study reach
which provided continuous information on
air temperature, precipitation, wind speed and
direction, solar radiation and relative humidity.
An automatic multi-parameter probe, inserted in
the stream bed, measured the water level every 30
minutes. Discharge measurements were recorded
in the study reach during and after the flood events
and for each month during the two hydrological
years from October 2007 to October 2009.
The SWAT model was calibrated using the
field measurement data from October 2007 to
October 2009 in order to establish the discharge
pattern throughout each of the two hydrological
years. The information was used to predict the
number of days without flow and infer the pattern
of ephemeral flow along different reaches and
tributaries throughout the catchment.
Methods
Establishing climatic patterns
and trends
Climatic information for the catchment was based
on temperature and precipitation data generated by
a long-term weather station at Evora (Instituto de
Meteorologia, http://www.meteo.pt/) and national
weather stations in Azaruja and Santa Susana
(Sistema Nacional de Informa¸ ao de Recursos
Hıdricos, http://snirh.pt/). The national stations
provided more than 60 years of temperature and
precipitation data. These were compared with
the 30-year average for the period 1961-1990,
in line with the minimum recommended time
period required to observe climatic trends. Climatic
trends were tested using the non-parametric
Mann-Kendall statistic test; time series patterns
were analysed using the so-called progressive and
retrograde series of the sequential Mann-Kendall
test in order to establish the likely beginning of
significant trends (Sneyers, 1990).
Results
Climatic trends
The pattern of annual variability in mean air
temperature at Evora during the period 1941-2006
showed a gradual increase when compared with
the average air temperature for 1961-1990 (Figure
22.2a). The Mann-Kendall statistical test confirmed
a trend of increasing mean air temperature. Both
progressive (dark grey line in Figure 22.2b) and
retrograde (light grey line) series began to increase
after 1977, but the upward trend only became
statistically significant (t
=
1.96,
=
0.05) in 2003,
following the intersection point between the two in
2002.
The annual variability of mean precipitation
during the period 1932-2007 indicates a gradual
decrease
Applying the Soil and Water
Assessment Tool
The 2009 version of SWAT (SWAT2009; ArcGIS
Version 9.3; http://swatmodel.tamu.edu/software)
was used to simulate a time series of discharge
hydrographs for 46 different locations in the
Pardiela catchment, selected randomly to represent
when
compared
with
the
average
