RIVER INPUTS
J. D. Milliman, College of William and Mary,
Gloucester, VA, USA
Copyright & 2001 Elsevier Ltd.
Introduction
Rivers represent the major link between land and the
ocean. Presently rivers annually discharge about
35 000 km
3 of freshwater and 20–22 Â 10
9 tonnes of
solid and dissolved sediment to the global ocean. The
freshwater discharge compensates for most of the net
evaporation loss over the ocean surface, groundwater and ice-melt discharge accounting for the remainder. As such, rivers can play a major role in
defining the physical, chemical, and geological
character of the estuaries and coastal areas into
which they flow.
Historically many oceanographers have considered the land–ocean boundary to lie at the mouth
of an estuary, and some view it as being at the head
of an estuary (or, said another way, at the mouth of
the river). A more holistic view of the land–ocean
interface, however, might include the river basins
that drain into the estuary. This rather unconventional view of the land–sea interface is particularly important when considering the impact of
short- and medium-term changes in land use and
climate, and how they may affect the coastal and
global ocean.
Uneven Global Database
A major hurdle in assessing and quantifying fluvial
discharge to the global ocean is the uneven quantity
and quality of river data. Because they are more
likely to be utilized for transportation, irrigation,
and damming, large rivers are generally better
documented than smaller rivers, even though, as will
be seen below, the many thousands of small rivers
collectively play an important role in the transfer of
terrigenous sediment to the global ocean. Moreover,
while the database for many North American and
European rivers spans 50 years or longer, many rivers
in Central and South America, Africa and Asia are
poorly documented, despite the fact that many of
these rivers have large water and sediment inputs and
are particularly susceptible to natural and anthropogenic changes.
The problem of data quality is magnified by the
fact that the available database spans the latter half
of the twentieth century, some of the data having
been collected 420–40 years ago, when flow patterns may have been considerably different to present-day patterns. In many cases, more recent data
can reflect anthropogenically influenced conditions,
often augmented by natural change. The Yellow
River in northern China, for example, is considered
to have one of the highest sediment loads in the
world, 1.1 billion tonnes y
À1
. In recent years, however, its load has averaged o100 million tons, in
response to drought and increased human removal of
river water. How, then, should the average sediment
load of the Yellow River be reported – as a long-term
average or in its presently reduced state?
Finally, mean discharge values for rivers cannot
reflect short-term events, nor do they necessarily reflect the flux for a given year. Floods (often related to
the El Nino ˜ /La Nin ˜ a events) can have particularly
large impacts on smaller and/or arid rivers, such that
mean discharges or sediment loads may have little
relevance to short-term values. Despite these caveats,
mean values can offer sedimentologists and geochemists considerable insight into the fluxes (and
fates) from land to the sea.
Because of the uneven database (and the often
difficult access to many of the data), only in recent
years has it been possible to gather a sufficient
quantity and diversity of data to permit a quantitative understanding of the factors controlling fluvial
fluxes to the ocean. Recent efforts by Meybeck and
Ragu (1996) and Milliman and Farnsworth (2002)
collectively have resulted in a database for about
1500 rivers, whose drainage basin areas collectively
represent about 85% of the land area emptying into
the global ocean. It is on this database that much of
the following discussion is based.
Quantity and Quality of Fluvial
Discharge
Freshwater Discharge
River discharge is a function of meteorological runoff (precipitation minus evaporation) and drainage
basin area. River basins with high runoff but small
drainage area (e.g., rivers draining Indonesia, the
Philippines, and Taiwan) can have discharges as
great as rivers with much larger basin areas but low
runoff (Table 1). In contrast, some rivers with low
457
J. D. Milliman, College of William and Mary,
Gloucester, VA, USA
Copyright & 2001 Elsevier Ltd.
Introduction
Rivers represent the major link between land and the
ocean. Presently rivers annually discharge about
35 000 km
3 of freshwater and 20–22 Â 10
9 tonnes of
solid and dissolved sediment to the global ocean. The
freshwater discharge compensates for most of the net
evaporation loss over the ocean surface, groundwater and ice-melt discharge accounting for the remainder. As such, rivers can play a major role in
defining the physical, chemical, and geological
character of the estuaries and coastal areas into
which they flow.
Historically many oceanographers have considered the land–ocean boundary to lie at the mouth
of an estuary, and some view it as being at the head
of an estuary (or, said another way, at the mouth of
the river). A more holistic view of the land–ocean
interface, however, might include the river basins
that drain into the estuary. This rather unconventional view of the land–sea interface is particularly important when considering the impact of
short- and medium-term changes in land use and
climate, and how they may affect the coastal and
global ocean.
Uneven Global Database
A major hurdle in assessing and quantifying fluvial
discharge to the global ocean is the uneven quantity
and quality of river data. Because they are more
likely to be utilized for transportation, irrigation,
and damming, large rivers are generally better
documented than smaller rivers, even though, as will
be seen below, the many thousands of small rivers
collectively play an important role in the transfer of
terrigenous sediment to the global ocean. Moreover,
while the database for many North American and
European rivers spans 50 years or longer, many rivers
in Central and South America, Africa and Asia are
poorly documented, despite the fact that many of
these rivers have large water and sediment inputs and
are particularly susceptible to natural and anthropogenic changes.
The problem of data quality is magnified by the
fact that the available database spans the latter half
of the twentieth century, some of the data having
been collected 420–40 years ago, when flow patterns may have been considerably different to present-day patterns. In many cases, more recent data
can reflect anthropogenically influenced conditions,
often augmented by natural change. The Yellow
River in northern China, for example, is considered
to have one of the highest sediment loads in the
world, 1.1 billion tonnes y
À1
. In recent years, however, its load has averaged o100 million tons, in
response to drought and increased human removal of
river water. How, then, should the average sediment
load of the Yellow River be reported – as a long-term
average or in its presently reduced state?
Finally, mean discharge values for rivers cannot
reflect short-term events, nor do they necessarily reflect the flux for a given year. Floods (often related to
the El Nino ˜ /La Nin ˜ a events) can have particularly
large impacts on smaller and/or arid rivers, such that
mean discharges or sediment loads may have little
relevance to short-term values. Despite these caveats,
mean values can offer sedimentologists and geochemists considerable insight into the fluxes (and
fates) from land to the sea.
Because of the uneven database (and the often
difficult access to many of the data), only in recent
years has it been possible to gather a sufficient
quantity and diversity of data to permit a quantitative understanding of the factors controlling fluvial
fluxes to the ocean. Recent efforts by Meybeck and
Ragu (1996) and Milliman and Farnsworth (2002)
collectively have resulted in a database for about
1500 rivers, whose drainage basin areas collectively
represent about 85% of the land area emptying into
the global ocean. It is on this database that much of
the following discussion is based.
Quantity and Quality of Fluvial
Discharge
Freshwater Discharge
River discharge is a function of meteorological runoff (precipitation minus evaporation) and drainage
basin area. River basins with high runoff but small
drainage area (e.g., rivers draining Indonesia, the
Philippines, and Taiwan) can have discharges as
great as rivers with much larger basin areas but low
runoff (Table 1). In contrast, some rivers with low
457