You are on page 1of 4

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/250125338

Rainfall Intensity-Kinetic Energy Relationships for Soil Loss Prediction1

Article  in  Soil Science Society of America Journal · January 1981


DOI: 10.2136/sssaj1981.03615995004500010033x

CITATIONS READS

162 909

1 author:

Peter Kinnell
University of Canberra
114 PUBLICATIONS   3,088 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Rainfall erosion processes and prediction View project

All content following this page was uploaded by Peter Kinnell on 18 August 2016.

The user has requested enhancement of the downloaded file.


i' lt t

\-
Reorinted
' from the Soil Science Society of Americalourru.l
Volume 45, no. l, January-Febniary f 981
6?? South Segoe Rd., Ma.liion, WI 5S71f USA

as a means of numerically describing the erosive Porver


of rainfall in equations designed to predict soil loss.
Because of teihnological liiritationi, devices capable
of measuring raindrop kinetic energy are not readily
available d.rop-size data from"'which raindrop
"riacan 6e calculated are not commonplace.
kinetic energy
FortunatelyJ rainfall kinetic energies calculated-from
drop-size.dlata have shown the rai-e of expenditure of
rairifail kinetic energy to be highly corielated with
rainfall intensity when raindrops are raveling at or
close to their tdrminal velocitiis (Hudson;8 Kinnell,
1973). Consequently, provided that the effects of wind
veloiity are i$nored,. intensity-kinetic energy relation-
ships provide-a means by which variations in rainfall
kinetii energy in time and space can be estimated.
Intensity-Energy Relationships
There are two forms of rainfall kinetic energy that
can be related to rainfall intensity. The first is the
rate of expenditure of rainfall kinetic energy (El.),
which has- units of energy /area/time. The se,cond is
the amount of rainfall kinetic energy expended per
unit quantity of rain (E*n). Eia has units of_ energy/
area/depth when rain'is eirpressed as a depth. Thus
RAINFALL INTENSITY-KINETIC ENERGY Eja is related to Ess by
RELATIONSHIPS FOR SOIL LOSS
PREDICTIONl
En1= cEssI'[ trl
where / is rainfall intensity (depth/time) and c is a
P.I.A. KrnNnrlz constant that adiusts for any diflerence that exists in
the units of time"used for / (..g., hours) and E11s (e.g.,
Absftact sec).
ilecause meteorological records frequently contain
Some of the inadequacies of equations used to describe
intensity data in the forrn of the amounts of rain
rainfall intensity-kinetic energy relationships :rre discussed in
connection with intensity-kinetic en€rgy relationshipa in Rho. recorded at various levels of rainfall intensity, the rela-
desia (Zimbabwe) and at Miami, florida. The relationships tionship between intensity and the kinetic -elergJ Pjr
between the kinetic energy per unit quantity of rain (8"1) and unit quantity of rain (the /-Epr relationtttip), is the
rainfall intensiry (I) at these two locations ar€ expr€ssed best one niost ofien used for estimating rainfall kinetic
by the equation energies for rainstorms.
Various types of mathematical equation have been
Ene = z (l'p6't't', used to describe /-Eno relationships. One has the form
where e is the base of natural logarithms and z, p, and h
are empirical constants. Eu=u*wlogro/ t2l
Ailditional Index Wotds: rainfall erosion. where u and w are empirical constants. This form of
equation is used to describe the /-Eno relationship from
Kinnell, P. I. A. 1981. Rainfall intensity-kinetic energy rela- which the R factor in the universal soil loss equation
tionships for soil loss prediction. Soil Sci. Soc. Am. J. 45:153- is calculated in the United States (Wischmeier 1959;
155.
Wischmeier and Smith, 1965, 1978).
Another has the form
Enn= c(b-aI-r) t3l
C on ERosroN by rainfall is a work process in the sense where a and b are empirical constants. This form of
D that work irivolves an expenditlre of energy, and I-Ene. relationship waf derived by Hudsons flom his
energy is required to detach soil material from the
soil surface and to transport this detached material in Rhodesia
observation that the /-Esa relationship in
away from an eroding area. Much of the energy re- (Zimbabwe) was essentially linear and could be de-
quired in the rainfall erosion process has been atri- scribed by
buted to the kinetic energy expended by raindrops Enn=bI-a. t4l
impacting with the soil surface (Hudson, l97l). As
a consequence of this, a number of indices based on Eq. [4] was also found to be valid in Miami, Florida
the kinetic energy of raindrop impact have been de- (Kinnell, 1973).
vised (Wischmeib, 1959; Hudson, 196l; Elwell, lgf78) Although Eq. t2] and t3l have been widely used
to estimate rainfall kinetic energies in the counuies
t C*trtbution from CSIRO, Canberra City, Australia. Re' where they were derived, they possess certain inade-
ceived ll Jan. 1980. Approved 27 Aug. 1980.
r Experiinental Officr-;r-, Sedimeht Transport Group, Division t N. W. Hudson. 1965. The influence of rainfall on the
of Soils, CSIRO Institute of Earth Resources, P.O. Box 639, mechanics of soil erosion, with particular reference to Southern
Canberra, A.C.T. 2601, Australia. Rhodesia. M.S. Thesis, Univ. of Cape Town, South Africa.

153
154
v
SOIL SCI. SOC. AM. J., vot. 45, l98l
gqac-rqr. example, Eq. tZ] tend.s to overestimate observed in /-Enn relationships, especially at
rainfall -Tor
kinetic energies a^t hign revels of rainfall in- levels of1. This, rogerher witli the i."tter'about t[e
high
tensity.- Becagse of ihis, a liirit of. 76 mm/hour (B /-Enn relationship (..g., Kinnell, lg73), is sufficient
inches/hour) has been imposed on the value'of use'd.
^I to mask nonlinearity at low levels of '.I when least-
to estimate rainfall kine-tic energies from Eq. t2] squares regression analysis is used to determine .I-Esp
(WischTeier and Smith, lg78). dn the orher'hahci, relationshins.
Eq. [3], .althoug,h tending td provide a better esti: Obviousiy, in the interests of obtaining more ac-
plte- at high levels of /, gives nelative values for rain- curate estimares of rainfall kinetic energ-y f-or the pur-
fall kinetic energy at valuei of /. For example, poles of predicting soil loss, it is necessiiy to desci-ibe
from the relationdhip -low
between I and Ess obtained'by I-Enn relationships by an equarion that combines the
Hudsons for Rhodesia, respectiv-eabilities of Eq. TZ] and t3] to describe
ERr 29.863 (14.287 I-t) I-Ena relationships at low anii high ievels of 1. An
= t5l equation that mCets this requiremEnt to some extent
(whe1e Esa is expressed in joule/m2 per mm and .I in has.been proposed by lVlcGregor and Mutchler (1977).
mm/hour), and negative va[ues of ^Esi occur when 1 ( I nls equatron rs
4.287 mniThour.
The inability of equations of the forms expressed Ena=z(l-pe-hl*Qe-tr\ t8l
-by.Eg.
t?] arid [3]^ to adequarely account ior the where e is the base of natural logarithms, and z, F,
variation in .Epa over all valubs of 1 results from the h, and-j are.empirical constant. -Eq. t8] was deiivedQ,

circumstances under which they were derived. Eq. flop diop-size- data measured at Fioliy'Springs, Mis-
[2], for example, was derived frilm drop-size data ob- sissippi,. and through the interaction bt ihe ierms p
luil.9.-by La*s and_ Parsons (1943) in Washington,
D.C. When one considers that Liws ahd parsons (f943)
e-ht and q ? .,,, proiluces (in the case of Holty Springi)
a sharply rising curve that peaks at a value of Esa -
measured raindrop sizes at relatively low rainfalt in- z when / is about 25 mm/hour (l inch/hour) before
tensities (2.^5_50.8 inmThour, 0.1-2.0 inchesThour), the decreasing- slowly as .I increases above 25 mm/hour.
inability of Eg.. [2J ro account for /-En,q, r6latioriships Although there-is no doubt that Eq. [8] provides a
1t high intensities is not surprising. On the other harid goog estimate- of the I-E*.e, relatiohshifi
-at Holly
[3]_, being_ derived from the relationship between Springs, a peak in Esa at low levels of 1-may not ba
-Fq and /, is biased toward the variation bf Ep,6 at
Ess evident at some geographic locations (FiS. l' and 2).
high rainfall intensities. The reason for this lies in In such cases, Eru-i when I +a, with'the result thjt
the fact -that Eas has an inherent, direct dependence Eq. t8l reduces ro
on.I. This is because (i)

Enn oc
l- 3 (NrDsKz2K)l t-LA-r t6l
l_K=l J Eq 2 t ERA=17.12t,,5.229 lo91g I

(where n is the number of drop-size categories, N is 1""t...

the number of drops having size-D and vei-ocity V, t is


.r' '..._

the time over which the sample is raken, and A is sam-


pling area), (ii)

Ia, |l_K=l
3 (N*D,*)l t-tA-t,
J
U] E
E
n.\
and (iii)_ the terminal velocity of a drop (Gunn and E

Kinzer, 1949) is approximately proportioial ro the o


square root of its size (D). The direct dependence of J
o
Ess on .I results in a high degr:ee of linEarity being tg.
t!
29.3 ( 1_0.281 e-0.018I )

Table l-Regreesion and conelation analyses for the relationehips


betweentt"T;11ffi1i,",fff":ili.i"antitvorrain
limits of doto scotter
(200 points
Residual Correlation )

Eq. Regression SS coefficient meon for 20 doto points

Miami, Florida (200 observations, 1.83-309 mrn/hour)


I2l ERA = L7.124 + 5.229log,o.I 3679 0.49
13l ERA = 30.132 (l-5.484.f-') 37849 Fig. f-The between the kinetic energy per unit
tel ERA = 29.3f [r-0.281 exp(-0.01841 360r 0.50 glaTt-ity-of-relationship
r_ain ({"i) ana rainfall intensity 1t1-'at'Miarni,
Rhodesia (50 observations, I 8. 5 - 228. 6 mr/hour) Florida-, lased on the E** values of Kinnell'(ig7gl. For thi
sake of clarity, o-nly_1hg mealls obtained per Z0 individual
I2l ERA = 9.705 + 9.2581og'o.I 537 0.61
.l?RA = 29.863ll-4.287 I-'l
pne values and the limits of the data enielope are shown.
t3l 542 0.61
ERA = 29.22 1L-0.894 exp( - 0.047741 418 0.7r
rndividual Eua values resulted from the apptication of the
tel
riry and vetoiiiy obtained by
relationship_terween-^d.rop
t{ne $orrle/m'per mm), / (mrn/hour). Constants for Eq. [2] and [9] were G'nn and Kinzer (1949)-to drop-size data,'obtained froni
determined by least sguares regression analysis on the f*o aata Con- qhologTpls of raindrops falinf through 0.l43ms of air
stants for Eq. [3] result from equations by Kinneil (lg?B) and i{udson.' during 10.5 sec. sample -1962).
periods (Mue[er,
\--
NOTES 155

7:pJ*-*t-:
(\l

o
2

E

E
/
/2 -
2t -t

xl
2t --F-
I
x
1o
lE Q 2: EP4=9;205+9.258 toqg I
o
'<2
t
Lll
Eq3lxuosotf t'
I
I
(r Epa = ZqeOf n- t 2S7 il I
I
I
t:
O meons for x or o

Eq 9 ' Enn=29.22 (i- 0.894 e-0"0t'77I I

100
t , mm,/hr
fig.- 3_ Tk _tslalfo.ryhip b€twem thc tinetic eDsgy per udt quantirV oI rein (E"r) and raiDfall tdte ity (4 fc thc data
of Hudlon.r Individual E"1 valuea were calculated f-ibm' raindrop 'sizes dhermined dy'i'iie flow petlct tectrniqui. iach rample con-
lilted o-f nfue llour panr (€.dl 77{ d) expoled in turn for'approximatcly 4 rec: s}mbols . ioa x uied to indicate'rhe in.
dividual Er^ valuea p-ertaining to eacb mean.- "t
Ene=z(t-Fe-n1. tgl across an electrical capacitor when the capacitor is being charged
via a resistor and, as a consequence of -this, proposed Eq. t9]
As can be seen from Fig. I Eq. tg] provides a better in a slightly different mathematical form. I would like to ihank
description of the I-En"e relationship at Miami than the unknown levierver who drew my attention to the similarity
can be achieved by Eq. t3]. In fact, the deviations of between the McGregor-Mutchler equation and the equation
originally proposed.
the observed values oT ER; from those estimared by
Eq. t3l {or I <.20_ mm/hour (20/o of the data) ii
so great that an index of correlition (Table l) can- Literature Cited
not be calculated for Eq. t3] at Miami. Statistically, l. Elwell, H. A. 1978. Modelling soil losses in Southern Africa.
there is little difference-between Eq. t2l and tgl 6t J. Agric. Eng. Res. 23:117-127.
Miami, but there is a tendency for Eq. L 2. Gunn, R., and G. D. Kinzer. 1949. The terminal velocity
iZlI to under- of fall waterdrops in stagnant air. J. Meteorol. 6:243-248.
estimate Esa at lo-r,v levels of / €ig. tj. 3. Hudson, N. W. 1961. An introduction to the mechaniscs of
The abilities of Eq. [2], [3], ;nd'[9] ro account soil erosion under conditions of subtropical rainfall. Proc.
for the variation in .Ena-af low ievels of / in Rhodesia Trans. Rhod. Sci. Assoc. 49:14-25.
are unknown. This is because there is a paucity of 4. Hudson, N. 1971. Soil conservation. B. T. Batsford Limited,
London.
data at I < 20 mm/hour (Fig. 2). However , for'I ) 5. Kinnell, P. I. A. 1973. The problem of assessing the erosive
20 mmThoul Eq. [9] provides a'better description of power of rainfall from meteorological obserr:ations. Soil
the ^I-Ep4 relationship j" Rhodesia (Fig. 2, Table l) Sci. Soc. Am. Proc. 37:617421.
than can be achieved by either nq. [Zl or Eq. tB]: 6. Laws-, J. O., and D. A.. Parsons. 1943. The relationship of
In gen_eral, th_e-refore, out of the th'ree iquatioris con- lai,ndrop size to intensity. Trans. Am. Geophys. Union 24:
452-460.
sidered, Eq. t9] provides the best esrimates of ERA 7. McGregor, K. C., and C. K. Murchler. 1977. Status of the
at Miami and in Rhodesia. When one considers that R factor in North Nlississippi. p. 135-142. In G. R. Foster
(ed.) Soil erosion: Prediction aird control. Proc. Nar. Soil
Fq. t9.] produces a curve that deviates only slightly
from llat prod.t.qq by the McGregor-Mutchler equa- Erosion Conf., Purdue Unir'., May 24-26, 1970, SCSA. An.
keny, Iowa.
-tion.(E_q.
8), Eq. [9] may possibly provide the means 8. Mueller, E. A. 1962. Raindrop distributions ar Miami, Flori-
by which reasonable estimat-es of rairifall kinetic energy da. Conrracr, DA-36-039 Sc-8?ZgO, Res. Rep. no. 9-B, Mer.
can be achieved in other geographic areas. Lab., Unir,. of Illinois.
9. Wischmeier, W. H. 1959. A rainfall erosion index for uni-
versal soil-loss equarion. Soil Sci. Soc. Am. proc. 23:246-249.
Acknowledgment 10. Wischmeier, W. H., and D. D. Smith. 1965. predicting rain-
fall-erosion losses from cropland east of the Rocky "Mo,rrr-
Originally,- because I was unaware of the McGregor-Mutchler tains. ARS Handbook no. 292. .\RS-USD.{
equation, I observed that the 1-,8*o relationships at il{iami and in ll. \Yischmeier, \V. H., and D. D. Smith. 1928. predicting rain-
Rhodesia were analogous to the transient behavior of voltage fall erosion losses. Handbook no. 537, USDA-SE.{.

View publication stats

You might also like