Tải bản đầy đủ - 0 (trang)
Chapter XVI. Surface Transports from Ecosystems

Chapter XVI. Surface Transports from Ecosystems

Tải bản đầy đủ - 0trang

424



XVI.



SURFACE T R A N S P O R T S FROM ECOSYSTEMS



lands, e.g., large wheat fields or pastures on the plains and above

timberline in the mountains, where no woody vegetation sticks up

high enough to brake the wind speed.

Maps of rates of snow drifting in the U.S.S.R. (Mikhel’ and

Rudneva, 1971) show that at the 0.05 level of probability, the amounts

during winter along the coasts of the Arctic Ocean are as great as 1000

m3 per running meter.. Inland, winter transport is about half as large

except on uplands, ”where the wind speed is higher, the frequency,

duration and intensity of blowing snow is greater.” In forested

regions, drifting is as little as 100 m3 per winter.

Measurements on the North American Great Plains showed that a

shelterbelt received about 50 kg of snow in a winter per square meter

of belt area (Sander, 1970). This extra water income is useful in

summer when transpiration from the trees is increased by heat

advection from the adjoining fields (the ”clothesline” effect) by about

35 kg m-2. In dry regions such winter transport of snow and subsequent buildup of soil moisture within tree belts can be vital to the

survival of the trees. The same mesoscale transfer of water operates in

the parkland belt of the Prairie Provinces of Canada. ”Trees and

shrubs improve their sites for future growth, by concentrating snowmelt supplies” (Laycock, 1972).

Loss of snow from dry-farmed lands where meltwater makes up a

critical part of the yearly recharge of the soil-moisture reservoir needs

to be prevented. Holding the snow on the fields gives the plants a

cushion of safety if summer rains turn out poorly. Benching and

stubble treatments of land in the northern Great Plains, intended to

hold rainwater for infiltration, also hold snow on the fields.

Soviet land-management methods in similar environments include

the use of grain stubble, rows of plant stalks (e.g., corn or sunflowers),

shelterbelts, and other devices to hold snow on the site where the

atmosphere delivered it. The drifting of snow is clearly an important

mode of off-site movement of water; L’vovich (1973, p . 174) estimates

the loss in Soviet forest-steppe and steppe zones as 16 km3 yr-’, and

advocates cutting it in half by shelterbelts and other measures. The 8

km3 so gained would be a substantial part of the additional water this

region needs to increase its agricultural yield.

* Since we are more used to visualizing the movement of water substance in channels,

it might be helpful to think of a small drainage basin, say of 10-km’ area. A flow rate of

0.2 m3 sec-’ in the outlet stream represents a yearly movement of 620 kg of water m-* of

basin area (or 620-mm depth). If this mass were transported out of the basin across a 5km-long side, the rate of flow would be 1250 m3 m-’ of border length, a number that can

be compared with the values cited above.



M O V E M E N T OF S N O W



425



Mountain ridges suffer the full force of the wind and are frequently

swept bare of snow. Such "schneearm" or "snow-poor'' ridges in the

Austrian Alps average 3 4 ° C colder, over the whole year even at

depths of a half meter (Aulitzky, 1961), and in winter as much as 15°C

colder in the top few centimeters compared with normal sites at the

same altitudes. They form a harsh environment for vegetation, having

little snow either to keep the soil warm or to protect the aerial parts of

the plants. As a result, these sites present a particular challenge to

foresters trying to restore a vegetation cover as one means of holding

snow at levels from which it would otherwise avalanche (Aulitzky,

1965).



Avalanching

Snow avaIanches so dramatically transport snow from its initial

lighting place on high slopes that their water-budget aspects might be

overlooked; yet many glaciers in mountain valleys are nourished by

avalanching snow, which prolongs the storage period of the mountain

water. At lower altitudes, valley-floor ecosystems show the evidence

of avalanche impact, as well as the massive input of water into their

annual budget. From heat-budget considerations during summer

ablation, Arai and Sekine (1973) estimate the inputs of snow necessary

to hold snow patches through the warm season at various altitudes

in Honshu. At 2 km altitude, an input of 8 tons of snow m-', mostly

from avalanching, will maintain a snow patch; at 1 km the input is 14

tons m-2.

In the high mountains of central Asia this vertical redistribution of

mass amounts to an off-site movement of about 0.1 of the snow stored

at high elevations (1-1.5 tons m-'), and a larger fraction of that in the

low-altitude reception zone (Abal'yan et al., 1971).

In contrast with the transport of snow by drifting, which is powered

by the kinetic energy of the wind, transport by avalanching is

powered by the potential energy of precipitation delivered from the

atmosphere to high-altitude slopes. As gravity is always in force,

forecasting avalanche descents requires data on the strength of the

snow mantle and observations are made to locate layers that are weak

in shearing strength or poorly bonded to the ground. Although

avalanches might occur at any time, they are especially likely to come

near the end of snowstorms that have loaded to failure the weak layers

holding the snow cover on a slope.

A classification of avalanches (Int. Assoc. Hydrol. Sci., 1973) points

up the off-site transport of mass that is our theme in this chapter. In



426



XVI.



S U R F A C E T R A N S P O R T S F R O M ECOSYSTEMS



the zone of origin, for example, an avalanche may start from a point

(”loose snow avalanche”) or from a line (”slab avalanche”). The

surface upon which sliding occurs may be at the ground or within the

snow mantle. The path may follow an open slope (”unconfined

avalanche”) or a gully (”channeled avalanche”); snow may move along

this path as a cloud of snow dust (”powder avalanche”), or along the

fracture plane, or along the ground surface. Finally, in the deposition

zone the snow may be clean, or contaminated by rock debris or tree

branches. Parallels between the debris-carrying capacity of avalanches

and that of the off-site movement of liquid water in sheets or channels

are obvious, and we can see other parallels between aspects of this

movement of snow, violent though it is, and the less dramatic off-site

flow of liquid water, both representing conversions of potential

energy bestowed by the ascent of condensing water vapor into the

middle troposphere.

GRAVITY-POWERED MOVEMENT OF LIQUID WATER



Gravity-powered transport of water in liquid state does not require

the steep gradients necessary in avalanche movement, but takes place

over gentle slopes if they are smooth enough that their surface friction

does not impede the flow. This phenomenon is not frequent; rain

reaches the earth’s surface during only a few percent of the hours of a

year, and usually at intensities less than the capacity of the soil to take

water in. Only at times are rainfall intensities great enough to exceed

the infiltration rate and to build up on the soil surface a layer of

detained water that will generate downslope flow. More frequently

cases occur in which the upper soil layers are saturated and horizontal

flow takes place near but just below the actual ground surface. The

fraction of total off-site flow that moves on top of the ground surface

depends on porosity, slope, and other factors. In such porous soils as

those of forests i t is almost never seen.

There is some advantage, nevertheless, in beginning with a rare

case, the storm Camille in 1969 in central Virginia, for which rainfall

conditions were discussed earlier. An eyewitness account follows:

. , . The lightning was brilliant and almost constant, it seemed,

accompanied by the sharpest claps of thunder I ever heard. But in

between these there was another sound-a roaring like forty jet

airplanes were stationary overhead. ”It must be that hurricane

picked up speed again,” my wife said. ”It must be blowing

dreadful outside.”



G R A V I T Y - P O W E R E D M O V E M E N T OF L I Q U I D W A T E R



427



Finally, at two o’clock, I got up to look out the window. The

inside window was up, but we’d closed the storm window to

keep the rain out. An inch of solid water was running down the

pane, so I couldn’t see out. I opened the window, and I got three

surprises. They were so sharp they were practically shocks. First,

the lightning was almost constant and very low, seeming to be

practically parallel to the ground. Second, the lawn outside, which

slopes down from the little swimming pool we have, was covered

with a solid sheet of clear water inches deep, rushing by the

house toward the creek. Third-and this was the biggest surprise

of all, though all of them were pretty big, let me tell you-was that

I had expected to find a screeching wind outside and it was

absolutely still. Not a breath. In the lightning flashes the tree

limbs hung down, the lower ones almost touching the ground

with the weight of the water, and they weren’t moving at all,

looking frozen in the eerie light. The roaring sound we heard was

the sound of water moving over the land (Kinkead, 1971).



Factors in Overland Off-Site Flow

This event was extraordinary. In rainstorms nearly every year,

though, it is possible to see detention films that build in the grass and

give rise to off-site flow on the ground surface. Since these events are

usually brief and field observations are few, experiments are done

with prepared plots subjected to simulated rainfall at high intensities.

Computer programs also can be developed to calculate the rapidly

changing fluxes that make up this seemingly simple phenomenon. For

example, the time required for an equilibrium flow rate from a slope

under uniform rain to become established has been studied and found

to be directly proportional to the length and roughness of the slope,

and inversely proportional to rainfall intensity (Morgali and Linsley,

1965). We shall return later to a more detailed consideration of these

factors, which are related to the depth of the film of water on the land

surface.

The discussion of detention storage in an earlier chapter shows that

not until the storage of water on the soil surface reached 5 mm did offsite movement of water begin. This did not come until 17 min after the

rain started. During the 100 min of rain, off-site movement from this

ecosystem occurred during only 12 min.

The detention layer built up from excess rainwater provides a

partitioning opportunity. From it water can move one way off-site, or

the other way down into the soil. Infiltration into the soil enjoys the

senior priority; off-site movement gets what is left. It is residual.



428



XVI.



S U R F A C E TRANSPORTS F R O M ECOSYSTEMS



There is controversy over the degree to which the transformation of

rainfall to flood fiow in a natural drainage basin, with its complicated

geometry and soil conditions, can be understood in a deterministic

way. Some hydrologists feel that present knowledge IS too small to let

us analyze the transformation deterministically, and that we must

resort to stochastic methods. Others are more confident, though

recognizing that the relations are seldom simple, being nonlinear and

aiso varying with time. The problem of water flow on slopes and in

channels, and of conceptualizing the processes by which water is

concentrate6 spatially, are questions of hydraulics more than of the

dynamics of water in soil-vegetation systems, however.

The rare of off-site flow of water q over the surface (per unit width!

depends on the depth of the detention film h , and on the impelling

force and the friction it must overcome. This force is the gradient- of

potentiai energy down the slope, and is opposed by the friction due to

vegetation a. One formulation of this relation is



q



=



ahn



in which n is an exponent approximately equal to 2, and a a parameter

expressing the gradient and roughness of the slope (Wooding, 1965).

Increasing the depth of the film h , which reduces the frictional drag

per unit mass of water, has a marked effect in increasing surface

run& 4,

The retarding effect of vegetation increases the time of concentration

of fiow. The longer time period gives the rain a chance to slack off;

and reduces the average delivey rate of water q at the bottom of the

sioge,



Overland Flow in a New Zealand Case Table 1, from a study in New

Zealand,, represents the way infiltration capacity, rainfall intensity,

and the detention storage of water impounded by vegetation frictior:

ail combine to generate a characteristic rate of off-site flow of water

The tabre shows runoff from a siope of 60-m length that results from

rainfall supplied at the intensity equal to probability 1.0 at Wellingtor..

(This intensity i decreases as the duration of the experiment increases.) The two kinds of grass in Table E illustrate different degrees

of friction in the film of moving water. The taller grass holds a greater

depth of water ponded in detention storage; its doye takes a longer

time to reach steady flow T,.Two representative values of infiltration

capacity, 1 and PO mm h y - ' ? represent differences in soii condition.

Four conditions of the vegetation-soil ecosystem are thus illustrated.

Because the mean intensity of rainfall i diminishes with the duration



TABLE I



E



Dominant Discharge from Small Areus with Three Types of Cover und Two Cupncities of Infiltration'



Vegetation



Infiltration

capacity f

(mm hr-l)



0



Buildup of detention

Time to reach steady

storage and surface

flow at downslope

Probable rainfall

intensity i during

runoff i - f

end of 60-m strip

(mm hr-I)

T , (hr)

period T , (mm hr-')



23



m



Runoff from 2-ha

area (md sec-')



z



4



i

l

_



Bermuda grass, dense uniform

cover 10-20-cm ht

Very short grass



' Source: Campbell (1967)



1



0.7



19



18



0.11



10



1.4



13



3



0.02



1

10



0.24

0.28



30

28



29

18



0.16

0.10



x)



C



430



XVI. S U R F A C E T R A N S P O R T S F R O M E C O S Y S T E M S



of the period, the slopes where movement of water is slow need a long

time to reach equilibrium flow, and over this period receive a

relatively low mean intensity of rainfall. Slopes from which the full

concentration of runoff occurs in a short time receive a higher mean

intensity. Water making its way slowly through the dense sod of

Bermuda grass takes 0.7 hr to get from the top of the slope to the

bottom and establish steady flow; water coming down the short-grass

slope takes 0.24 hr. These periods are called times of concentration T,.

At 0.7 hr, the entire area of the Bermuda grass slope is contributing

water that reaches the downslope end. This duration is long enough to

allow the rainfall intensity to slacken a bit. This aspect of duration,

which might not be obvious at first glance, derives from the tendency

of rainfall intensity to decrease as its duration increases, which we

considered in Chapter 111. The mean intensity of the once-a-year rain

at Wellington during the 0.7-hr concentration time associated with the

Bermuda grass slope, is 19 mm hr-'. This is a much lower rate than the

mean intensity, 30 mm hr-', in 0.24 hr, the time relevant to the shortgrass slope. A bare slope or a parking lot, which would have a still

shorter concentration time, would have to handle a still higher

intensity rainfall.

The other variable illustrated in Table I is infiltration capacity f ,

shown for each slope for two selected soil conditions, 1 and 10 mm

hr-l. In determining the rate at which detention storage builds up, the

respective infiltration rate f is subtracted from the rainfall rate i. This

excess water i - f goes first to build up detention storage (including

filling microdepressiors), then into surface runoff. In the low-infiltration case of the Bermuda grass slope, the water excess is 19 - 1 = 18

mm hr-'. This is six times as much as on the same slope under highinfiltration conditions.

Runoff rates are high if vegetation is short and offers little frictional

resistance to overland flow, and also if infiltration capacity is low.

Runoff from the four soil-vegetation systems in this once-a-year rain

ranges from 0.02 to 0.16 m3 sec-', that is, over nearly one order of

magnitude.



Changes in Soil-Vegetation Systems: New Zealand A later study in

New Zealand identifies how the critical parameters of infiltration

capacity and detention storage are changed as a result of practices that

are transforming the agriculture of the country. Higher applications of

phosphate fertilizers, often made by airplane to hill-country farms,

have increased the growth and density of grass over millions of

hectares (see Plate 34).



G R A V I T Y - P O W E R E D M O V E M E N T OF L I Q U I D W A T E R



431



Plate 34. Hill-country agriculture on the North Island of New Zealand, with aerial

fertilization (Te Kuiti, February 1966).



In one experimental area, where the grass production rate was

tripled (Toebes et al., 1968), the hydrologic results of the greater plant

vitality were evident as increases in infiitration capacity and surface

detention. A greater density of plant stems and leaves slows the flow

of surface water; greater plant activity dried out the root zone,

increasing its capacity to take in water from the next rain. These

changes result in less off-site fiow. Moderate rains that formerly

caused off-site flow on untreated slopes were completely accommodated on the treated slopes, and produced no surface runoff.



Changes in Soil-Vegetation Systems: Coshocton Experimental Site Because

deeper-rooted grass keeps the soil in better condition and dries it out

more rapidly, water from each rainstorm finds greater opportunity to

infiltrate, i.e., less water piles up on the surface to move off-site. Table

11, surface runoff from summer rains at Coshocton, Ohio demonstrates

that more vigorous plant growth means less off-site movement of

water. Grass yield increases 5 times, water yield decreases to$ .

Conversely, infiitration into an already moist soil is slower than into

a dry one for the reasons discussed in an earlier chapter. In a

protracted rainy period infiltration is low even at the beginning of

each burst of rainfall, and surface runoff from the burst is correspondingly increased. Surface runoff was measured over thousands of plot



432



XVI.



SURFACE T R A N S P O R T S FROM ECOSYSTEMS



TABLE I1



Surface Runoff in Summer from Two Kinds of Grass in Lysimeters at

Coshocton, Ohio“



Parameter



Poverty grass

(shallow-rooted)



Annual biomass yield (g m-’)



110



Summer runoff

1959

1960

1961

1962

1963

Mean of five summers

a



Birdsfoot trefoil

(deep-rooted)

570



2

10

8

11

20

8



Units: mm. Source: Dreibelbis and Amerman (1964).



years at 35 soil erosion experiment stations in the United States and

found correlated to rainfall energy and 30-min intensity. It was also

found to be correlated with rainfall in the preceding 24 hr, which

influences the level of soil moisture and hence the rate of infiltration

(Nelson, 1958) and runoff.



Surface Runoff from Some of the Ecosystems in a Drainage Basin

Soil bodies already saturated at the beginning of a rain will not take

in any water by molecular attraction, and detention storage will almost

immediately start to form on the surface. Off-site runoff soon follows.

It might be a flow on top of the saturated soil or a pressure wave

forcing water that has filled the large pores of the soil to start moving

out the downslope end of the soil body.

Saturated soil bodies are common near streams, in shallow soils,

and at the foot of slopes. Their rapid response to a burst of rain that

elsewhere in a drainage basin might infiltrate into the soil often

supports the initial rise in the nearby stream channel. The conversion

of rainfall to quick runoff is virtually complete, and the stream rise can

be substantial even if no more than 0.1 of the area of the basin is

contributing (Kirkby, 1969). Horton (1937) identified and studied

partial-area rises” in rivers, and presents a case study of the Muskingum River in June 1929. In later generalizations of the excess-rainfall

concept for application to large areas, it was commonly assumed that

all parts of a drainage area contribute equally, but more recently the

I,



433



G R A V I T Y - P O W E R E D M O V E M E N T O F L I Q U I D WATER



25



0



0



5



15



50

10



15



20



FEET



METERS



Fig. XVI-1. Areas of near-stream land that contribute storm flow in intense summer

storms (black) and in autumn storms under generally wet conditions (from Dunne and

Black, 1970).



partial-area idea has had a revival in explaining the time distribution

of storm flow* as a spatially integrated flux from many ecosystems.

As a storm continues, the number of ecosystems with full storages

of interception, detention, and soil-moisture capacity increases. Soon

all are contributing storm runoff, and channel flow builds up to a

peak.

A seasonal variation also occurs in the size of the contributing area.

Storms in fall and winter deliver rain to a basin in which a large area

is wet and ready to convert rain to storm flow; in summer the area

ready to make this conversion is much smaller (Fig. XVI-1) (Dunne

and Black, 1970).

* ”Storm flow” refers to immediate off-site flow, whether it travels entirely on the

ground surface, in the top layers of the soil, or both.



434



XVI.



SURFACE T R A N S P O R T S FROM ECOSYSTEMS



The idea that all the ecosystems to which water is delivered by a

rainstorm are not necessarily equally ready to take it in expresses the

diversity of nature, which is manifested in the contrasts among

ecosystems with respect to soil depth and texture, vegetation density,

and other properties of hydrologic significance. ”The partial area

concept describes, qualitatively, what one can observe in the field.

Quantitatively, one cannot say as much. To date, no successful

quantitative procedure has been developed to predict watershed

runoff on this basis” (Engman, 1974, p. 514).



Flow Near the Surface of the Ground

In contrast with off-site flow across the land surface of experimental

plots or thin bodies of saturated soils, some ecosystems almost never

experience overland runoff. This behavior reflects ecosystem characteristics or situation.



In Forest Ecosystems The effect of different soil-vegetation systems

on storm-flow events is shown in Table I11 from measurements in the

dissected land of southwestern Wisconsin. The virtual absence of

storm flow on forested slopes, even when they are very steep,

indicates the hydrologic asset that the forest floor represents. Its open

structure can accommodate large inputs of water. Much of this moves

off-site within the soil, but more slowly than if it moved over the

surface, and does not produce an initial steep rise of the flood wave.*

This near-surface flow makes the main contribution to the flood wave

itself, however. In the mountains of the northeastern United States,

”the principal source of flood runoff delivered to the channels in

forested watersheds is subsurface flow that moves rapidly downslope

through the permeable forest floor and topsoil” (Lull and Reinhart,

1972, p. 83).

A good deal of folklore on forest effects on floods has accumulated

during the past century, some of it exaggerating their undoubted

benefits to runoff. Sartz (1969) discusses some of these exaggerations,

many of them based on old research and crude measuring methods.

Relation to Slope Porous soil and high infiltration capacity of

forested slopes permit little rainwater to move over the surface so

there is little action of water on the surface soil. As a result, slopes

remain uneroded, and unreduced in steepness. The coincidence of

*This condition exists even in the heavy rainfall of the southern Appalachians.

Surface flow has never been seen to cross some siopes of the Coweeta Experimental

Forest in more than 20 years of observations (Helvey et al., 1972).



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

Chapter XVI. Surface Transports from Ecosystems

Tải bản đầy đủ ngay(0 tr)

×