FIGURE 4.11 Comparison of the Lag-1

FIGURE 4.11 Comparison of the Lag-1 autocorrelation for 10 stations on the Lower Mekong River, compared with 624
global stations (data set compiled by Murray Peel).

The large rivers
of tropical Africa, such as the Congo, have
relatively modest flood regimes in terms of unitdischarge, which is attributable to a combination
of low relief and less extreme tropical rainfall
climates.
By comparison with many other rivers, the
size of the flood pulse in the Mekong is extremely
predictable. Campbell (Chapter 16) noted that
the Cv of the annual flow of the Mekong at
Chiang Saen is only 0.2, while the worldwide
average for rivers with catchments larger than
105 km2 is 0.33 (McMahon et al., 1992). Further
downstream, at Pakse, the site closest to the
Cambodian floodplain, the Cv has dropped to
0.16, less than half the global average.
Even though the flood size is extremely consistent
from year to year, this does not mean
that the size of the flood in one year is a good
guide to the size of the flood in the next year.
The Lag-1 autocorrelation explores the relationship
between floods in successive years. In
catchments with high positive correlation, of
0.6 for example, this would mean that the magnitude
of the flood in one year was quite a
strong predictor of the size of the flood in the
next year. In general, the floods on the Mekong
are only weakly, positively correlated with the
floods of the year before (explaining between
Hình 4.11 So sánh các Lag-1 tự tương đối với 10 trạm trên sông Mêkông, so với 624
trạm toàn cầu (bộ dữ liệu được biên soạn bởi Murray Peel).

10% and 20% of the variation) (Fig. 4.11). This
means that the size of floods from year to year
is essentially independent of each other, which
is typical of rivers of this size.
4. HISTORICAL CHANGES IN
HYDROLOGY
Decadal variation in the hydrology of the
Mekong River provides context for exploring
the possible effects of both human impacts on
hydrology, and climate change. The Mekong
experiences quasi-periodic discharge fluctuations
at an interdecadal scale. This fluctuation
is visible in the annual Mekong dry-season
flows at Vientiane (Fig. 4.12), particularly the
substantial decrease in dry-season flows
between the 1940s and 1950s.
Human impacts on hydrology can be classified
as direct and indirect. The major direct
impact on water volumes in the Mekong is diversion
for irrigation. Indirect effects are caused by
dams and changes in land use, in particular the
conversion of forest to agriculture. These impacts

can alter the gross volume of water in the river,
as well as the timing and duration of flows. There
has been much speculation about the effect of
human impacts on flow regimes, but little investigation
of the evidence, as described below.
4.1. Irrigation
Removal of water for irrigation is the largest
direct hydrological impact on the Mekong
River. Simulations of a 20-year flow period for
the Mekong river basin indicates irrigation
water requirements of 13.4 km3 year, which corresponds
to a 2.1% and 2.3% decrease in the
mean annual streamflow at the outlet (Haddeland
et al., 2006). Half of the diverted water is
estimated to be lost via evapotranspiration,
and half returned to the river (Jackson et al.,
2001). While this is a substantial volume of
water, when compared with irrigation water
demand from other large rivers, this is a modest
diversion. For example, 37% of the total volume
of the Colorado River in North America is
diverted for agriculture (Haddeland et al., 2006).

Although the volume of water diverted for
irrigation is modest, it is important to note that
this diversion occurs in the dry season, when
the relative effect is greatest. For example, in
the delta at Phnom Penh in February, March,
and April, it is estimated that 60%, 45%, and
40% (respectively) of the flow is abstracted for
irrigation (MRC, 2003). It is worth noting that
the majority of dams planned for the Mekong
Basin are designed for hydropower generation
rather than for water extraction. The effect of
these dams will be to increase dry-season flows
(Podger et al., 2004) which could compensate
for increases in dry-season irrigation extractions.
4.2. Effects of Deforestation
Forest degradation in the Mekong Basin has,
according to Giril et al. (2001), been occurring at
an unprecedented rate and scale, particularly
from the 1960s onwards (Table 4.5). On the
Korat Plateau in Thailand, which includes the
Mun and Chi tributary systems, forest cover
was reduced from 42% in 1961 to 13% in 1993

(MRC, 2005). Furthermore, logging pressure on
the forests of Lao PDR, Cambodia, and Burma
was intensified after 1989, when Thailand introduced
a logging ban within natural forests, and
consequently sought increased imports from its
neighbors.
Two potential hydrological impacts of deforestation
might be distinguished:
1. Total water yield may be increased as annual
evapotranspiration decreases, and
2. Seasonal distribution of flows may be modified
as flood runoff increases and dry-season