CHAPTER III

PRECIPITATION
(With emphasis on Rain)

LEARNING OUTCOMES:

At the end of the chapter, students should be able to:

1. explain basic terms about precipitation

2. list mechanism related to occurrence of precipitation

3. name precipitation kinds and forms

4. estimate missing data

5. calculate average precipitation

3.1 FORMATION OF PRECIPITATION

► Precipitation is that part of atmospheric moisture that falls on the earth

surface. It is the primary input into the hydrologic cycle. It could occur in
various forms such as rain, snow, sleet, or hail.

► It is generally agreed that the initial process in the formation of

precipitation is the condensation or sublimation of atmospheric moisture.
This transformation may be brought about by one or more of the following
causes: (1) adiabatic cooling; (2) mixing of air masses of varying
temperature; (3) radiation cooling and (4) contact cooling. The more
important of these causes are adiabatic cooling and mixing of air masses of
different temperatures. The two other causes rarely produce significant
amounts of precipitation.

► The condensation into water droplets and sublimation into ice crystals of
atmospheric water vapor generally occur around condensation or
sublimation nuclei. The principal types of these nuclei are particles of
ocean salt, products of combustion and oxides of nitrogen. These nuclei
are usually less than 1 micron in diameter. Usually, it only takes a few
seconds to produce water droplets of 10 microns which is the average size
of water droplets in clouds. Such droplets, however, weigh so little that an
upward air movement of less than 0.15 m/min is sufficient to keep them

69
 from falling into the earth’s surface. Ice crystals of the same diameter are
even lighter because of their smaller specific weights and can be supported
by even lower air velocities.

► Even in the absence of upward wind draft, very small water droplets or ice
crystals falling from clouds evaporate while passing through unsaturated
atmospheric layers. For example, it has been reported that water droplets
with diameters of as much as 200 microns will evaporate in a 3.3 m thick
atmospheric layer having a relative humidity of 90 percent. Thus, most ice
crystals and water droplets formed by condensation and sublimation never
reach the earth’s surface.

► Raindrops reaching the earth’s surface are generally over 500 microns (0.5
mm) in diameter. A small raindrop is usually about 3 mm in diameter. It
has been estimated that it will take about a day for the condensation
process alone to form such a drop. As a result, condensation alone can not
be expected to produce any significant amount of precipitation.

► Two other processes are generally regarded as capable of supporting the

growth of water droplets or ice crystals formed by condensation and
sublimation. These are the coalescence and ice crystals processes. Implied
in the coalescence process is the fusion or coalescence of water droplets as
a result of their coming in contact through the action on them of air
movement and gravitational pull. Differences in size and wind draft will
cause differences in drop velocities and, hence, increase the probabilities of
collision and union of drops to form raindrops of from 0.5 to 4.0 mm in
diameter. The coalescence process may also be a result from electrostatic
attraction of cloud elements (including water droplets) of different
electrical charges.

► The growth of ice crystals or water droplets through the ice crystals process
is caused primarily by the difference in the vapor pressure around water
droplets and ice crystals. Ice crystals and water droplets are known to co-
exist in the atmosphere at sub-freezing temperatures. Under such
condition, the vapor pressure around a water droplet is higher than that
over an ice crystal. This difference in pressure serves as the driving force
for condensation of moisture over the ice crystal. This growth of the ice
crystals also favors coalescence with other ice crystals or water droplets.

► The importance of understanding the process of rain formation lies in the

possibility of artificially inducing precipitation.

70
3.2 FORMS OF PRECIPITATION

► Any product of condensation of atmospheric water vapor formed in the free

air or at earth’s surface is called a hydrometeor. Only those hydrometeors
falling into the earth’s surface, termed precipitation, are of particular
interest to agricultural engineers as far as water resources management and
development is concerned. In the Philippines, there are only two forms of
precipitation commonly observed, as follows:

1. Drizzle – sometimes called mist, consists of tiny liquid water droplets,

usually with diameters between 0.1 and 0.5 mm (0.004 and 0.02 in),
with such slow settling rates that they occasionally appear to float. This
usually falls from low stratus and rarely exceeds 1 mm/hr (0.04 in/hr) in
intensity.
2. Rain – consists of liquid water drops mostly larger than 0.5 mm (0.02
in) in diameter. Rainfall usually refers to amounts of liquid
precipitation. In the United States, rain is reported in three intensities:

Light : For rates of fall up to 0.10 in/hr (2.5 mm/hr)

inclusive.
Moderate : From 0.11 to 0.30 in/hr (2.8 to 7.6 mm/hr)
Heavy : Over 0.30 in/hr (7.6 mm/hr)

3.3 TYPES OF PRECIPITATION

► Adiabatic or dynamic cooling is the primary cause of condensation and

hence, is responsible for most precipitation. Adiabatic cooling results
mainly from expansion of air mass which results from a decrease in
atmospheric pressure. This is triggered primarily by the upward or vertical
motion of air masses. As a result, storm or precipitation events are usually
classified according to the conditions that cause this upward motion. Three
general types of storms are identified in this respect, as follows:

1. Convective Precipitation. Typical of the tropics, this type of

precipitation results from the differential heating of air masses near the
ground surface resulting in the upward movement of warmer air masses.
A heated air mass moving upward is cooled down by the surrounding
air and by the expansion process. Cooling brings about condensation
which may develop into precipitation. This type of precipitation may be
in the form of light showers to cloudburst or thunderstorm depending on
the temperature and moisture conditions. Convective precipitation are
usually highly localized and of short duration.

71
2. Orographic Precipitation. This type of precipitation is influenced by
topography. Moist air masses moving over mountain barriers are
cooled, resulting in condensation and precipitation. Such air masses get
dry and warm as they move downslope on the leeward side of the
mountain barriers. Orographic types of rainfall are usually of very low
intensity.

 The influence of this type of precipitation on the climate of many areas

in the Philippines is very significant considering that the Philippines is
mountainous and made up of a chain of islands.

3. Cyclonic Precipitation. Cyclonic precipitation are associated with

the movement of air masses due to differences in barometric pressure.
This type of precipitation may further be classified into two general
types: frontal precipitation and non-frontal precipitation. The non-
frontal precipitation is triggered by the presence of a low pressure into
which air flow converges resulting in the lifting and subsequent cooling
of air masses. Frontal storms (which produces frontal precipitation)
result when two air masses of different characteristics such as warm,
moist air mass and cold air mass meet. Precipitation is usually
produced at the frontal or colliding boundaries of these air masses.
Significant rainfall (warm front precipitation) of longer duration and
wider areal coverage usually results when warm air masses move
upward over cold air masses. Less intense rainfall results when an
advancing cold front pushes back warm, moist air masses (cold front
precipitation).

 In most parts of the Philippines, more than 80 percent of the total

rainfall are attributable to cyclonic precipitation. The more common
weather phenomena causing cyclonic type of rainfall include
tropical cyclones, barometric lows during the summer months
inducing the moist southwest monsoons, the movements of the
Intertropical Convergence Zone (ITCZ) and the southward
movement of cold, high pressure cells from temperate Asia during
the winter months.

 Tropical cyclones are destructive weather disturbances that usually

develop near the equator (5 deg north or south). They are
characterized by low pressure centers with great air masses spiraling
towards them. They move to higher latitudes at about 10 to 50 km
per hour in a northwesterly direction. The general atmospheric
conditions during the period from July through September favor the
drawing into the Philippines of cyclones that develop in the Pacific
Ocean. On the average, about 35 to 45 percent of such cyclones

72
 reach the eastern coasts of the country. After September, changing
weather conditions cause the cyclones to move westward toward the
southern parts of the Philippines. Also, the chances of a cyclone that
develop over the Pacific Ocean reaching the Philippines
progressively diminish (10 to 25 percent in October and November,
and less than 5 percent in March). On the average, the country is
visited by about 15 to 22 tropical cyclones each year. Each cyclone
usually deposits considerable rains along its wide area of influence.

 The Intertropical Convergence Zone (ITCZ) is the zone of

convergence of the trade winds coming from the north and southern
hemispheres. As a region of convergence of two air masses, it is
characterized by widespread cloudiness, precipitation and moderate
to strong surface winds. The ITCZ is not stationary near the equator
but moves with respect to the orientation of the earth’s axis as it
orbits the sun. During the summer months in the northern
hemisphere, it is usually found north of the Philippines. In winter
months when high pressure cold cells hover over the Asiatic
mainland, it migrates south of the equator. Its influence on
Philippine weather is felt mainly during the period from May to
October.

 The temperature variations over the Asiatic continent as a result of

changing inclination of the earth’s axis cause changing air streams
which, in turn, significantly influence Philippine weather. During
the winter months from December through March, the cold air mass
over mainland Asia send strong northeasterly winds over the
Philippines causing precipitation over the eastern coastal areas of the
country. In the summer months, the Asiatic mainland becomes
warmer than its surrounding oceans and a low pressure area develops
over the continent. This causes the flow of moist, warm winds
coming from the southwest. Originating from the Indian Ocean,
these southwest monsoons are of considerable depth (over 10,000 m)
and width and persists for a long period (May through October).
They induce prolonged and widespread rains in Luzon and the
Visayas.

3.4 ARTIFICIALLY INDUCED PRECIPITATION

► Weather modification, sometimes referred to as weather control, is the

general term for efforts to alter artificially the natural meteorological
phenomena of the atmosphere. Attempts to increase or decrease
precipitation, suppress hail and lighting, mitigate tropical cyclones,

73
 dissipate fog, prevent frost, alter radiation balance, etc., are all included
under weather modification. Cloud modification, or cloud seeding, is one
type of weather modification, and usually has as its goal either dissipation
of the cloud or stimulation of precipitation.

► It was demonstrated in 1946 that dry ice can cause precipitation in a cloud
containing supercooled water droplets. This discovery soon led to further
discoveries that certain salts, notably silver iodide, can also induce
precipitation. Both dry ice and silver iodide, the two most commonly used
seeding agents, act as freezing nuclei in supercooled clouds. Seeding
clouds with dry ice requires delivery into the aircraft, balloons, or rockets.
Silver iodide, which is most effective when heated to vaporization, may be
delivered into the cloud by either airborne or ground-based generators but
has the disadvantage that its effectiveness is reduced by exposure to
sunlight, the number of effective particles decreasing by a factor of about
10 for every hours of exposure. Nevertheless, the low operation cost of
ground-based generators has made this method the most commonly used
for augmentation of precipitation.

► The effectiveness of cloud seeding depends on many factors such as height

of cloud base and top, cloud temperature, difference between density inside
the cloud and that outside (buoyancy), updraft velocity distribution, amount
and concentration of liquid water in the cloud, number and distribution of
natural freezing or condensation nuclei, the number of artificial nuclei
added, and where they are introduced into the cloud.

3.5 MEASUREMENT OF RAINFALL

3.5.1 GAGING RAINFALL

 The purpose of the rain gage is to measure the depth and intensity of
rain falling on a flat surface. There are so many problems of
measurements with gages which include effects of topography and
nearby vegetation as well as the design of the gage itself. Rain gages
are generally vertical, cylindrical containers with top openings 203
mm in diameter. A funnel-shaped hood is inserted to minimize
evaporation losses.

 Rain gages may be classified as recording or non-recording. Non-

recording rain gages are economical, require servicing only after
rains, and are relatively free of maintenance. An example of a non-
recording rain gage is the standard rain gage (Fig. 3.1) which is a
cylindrical tank having a circular receiver 8 inches in diameter. For

74
 each of measurement of small precipitation amounts, the rainfall is
funneled from the receiver into small inner cylindrical measuring
tube which has an area that is one-tenth that of the receiver. This
provides a magnification of 10 times the depth of the water and
makes it possible to measure to the nearest 0.25 mm.

Fig. 3.1. The standard 8-inch rain gage.

 Many practical applications dictate that it is important to measure

not only the total rainfall or rainfall depth but also the rainfall
intensity or rate of rainfall. Recording rain gages similar in
dimensions to those employed for measuring rainfall depth are used
for this purpose. Recording rain gages may be of the weighing type
which record on a chart the cumulative amounts of rainfall with
time, or may be of the tipping bucket type (Fig. 3.2) where two
buckets alternatively collect then empties 0.01 inch depth of rainfall.

75
 Fig. 3.2. The tipping bucket rain gage.

3.5.2 ERRORS IN MEASUREMENT

 Many errors in measurement result from carelessness in handling the

equipment and in analyzing data. Errors characteristic of the non-
recording rain gage of the U.S. National Weather Service include
water creeping up on the measuring stick, evaporation, leaks in the
funnel or can, and denting of the cans. The volume of water
displaced by the measuring stick is about 2 percent and may be taken
as the correction for evaporation.

 Another class of errors is due to obstructions such as trees, buildings,

and uneven topography. These errors can be minimized by proper
location of the rain gages. The gages are normally placed with the
opening about 760 mm above the ground surface. They should be
located so as to minimize turbulence in the wind passing across the
gage. A practical rule is to have a clearance of 45 degrees from the
vertical center line through the gage, but a safer rule is to be sure that
the distance from the obstruction to the gage is equal to at least two
times the height of the obstruction.

 The wind velocity also affects the amount of water caught. A wind
speed of 16 km/hr would decrease the rainfall caught by the gage by
about 17%, but at 48 km/hr the deficit is increased to about 60%.
Whenever possible, the gage should be located on level ground as

76
 the upward or downward wind movement often found on uneven
topography may easily affect the amount of precipitation caught.

3.5.3 THE GAGING NETWORK IN THE PHILIPPINES

 Rainfall is the first climatological parameter extensively monitored

throughout the Philippines. Many agencies (PAGASA, the National
Irrigation Administration, the Bureau of Forest Development and the
Bureau of Plant Industry) operate a nationwide network of gaging
stations. State colleges and universities and private entities also
maintain one or more rain gages in their various localities.

 There are about 100 gaging stations in the country with long (25
years or more) years of daily rainfall records. Most of these,
however, are located along the coastal areas as they were established
primarily for aviation, transport and non-agricultural purposes.
There are about twice as many stations having anywhere from 8 to
24 years of records. Most of these are established mostly for
agricultural, forestry and research and development purposes.
Unfortunately, there is no single agency maintaining a complete
record of most of these stations.

 The present gaging network density is still farm from desirable.

Even if it is assumed that there are at present about 300 gaging
stations throughout the country and that these are uniformly spaced,
the resulting network density is still only one rain gage for about
30,000 hectares. In terms of sampling to total area represented
ratio, this is equivalent to about 1.1 x 10-10 as the area sampled by a
standard rain gage is only about 0.032 m2.

3.6 ANALYSIS OF PRECIPITATION DATA

3.6.1 COMPLETING AND ADJUSTING PRECIPITATION DATA

 In not a few cases, the precipitation data records for a given gaging
station have missing observations or cannot be accepted at face
value. In some cases, the gaging station locations had been changed
during the period for which the series of records is available. For
proper analysis and interpretation, the records series must be made
complete and some data entries adjusted to reflect changes in gage
sites.

77
 Estimation of Missing Precipitation Data. For longer term totals
such as seasonal and annual rainfall magnitudes, the normal-ratio
method may be used. In this method, the three nearest stations with
long years of records are chosen as index stations. If the normal
precipitation in each of these index stations is within 10 percent of
that for the station with the missing record, a simple arithmetic
average of the precipitation totals of the index stations for the time
period in question will suffice. Otherwise, the normal precipitation
values at the index station are weighted as follows:

Px = 1/3 {Nx (Pa/Na) + Nx (Pb/Nb) + (Nx (Pc/Nc)}

(3.1)

where: Na, Nb, Nc = the normal precipitation totals for the index
stations A, B and C for the desired time
periods.
P a, Pb, Pc = precipitation totals for the desired time period
for the three index stations.
N x = normal precipitation at station X
for the desired time periods, and
Px = estimated precipitation total for station X
(with missing records)

Missing rainfall records for shorter time periods may be

estimated using a regression model of the form:
n
Rm = Bo +  Bj Rj + Bn+1 Rp + Bn+2 Rs + 
(3.2)
j =1

where: Rm = estimate of missing rainfall for a particular period.

Bo = intercept of the regression line.
B’s = regression coefficients.
Rj = rainfall during a particular period in the jth
neighboring or index equation.
Rp = rainfall during the preceding period in the station
with missing observation.
Rs = rainfall during the succeeding period in the
station with missing observation.

The use of the regression equation is based on the assumption that

the magnitude of rainfall in a given station is related to the
occurrence or non-occurrence of rainfall in the adjacent stations with
the same climate type and that the magnitude of precipitation in that
station during the period
78
 with missing data is related to the magnitudes of precipitation in that
station during the preceding and succeeding periods.

 Adjustments of Rainfall Data. The sites of many rain-gaging stations

in the country had been changed. Such changes are not usually
disclosed in the published reports. Double-mass analysis is usually
used to test for the consistency of records at a given site. This method
compares the annual or seasonal rainfall totals at the station in question
with mean values of a group of surrounding stations.

Figure 3.3, for example, shows a hypothetical double mass

analyses curve for a given station X. The annual rainfall totals for
station X were plotted against the mean values of the precipitation
totals of index or neighboring stations. A break in the slope of the
curve in 1946 indicates that station X was moved in 1946.

A break in the slope of the curve would indicate a change in

precipitation regime at station X. Such a change can be attributed
to change in location or exposure of the raingage since a change due
to meteorological factors will also be reflected in the index stations
and will not cause a change in the slope of the curve.

To adjust the records prior to 1946 to those corresponding to

the new site, the annual precipitation totals prior to 1946 should be
multiplied by the ratio of the two curves. That is, all annual totals
before 1946 must be multiplied by the ratio 1.05/1.44.

3.6.2 INTENSITY, DURATION AND FREQUENCY OF RAINFALL

 One of the most important rainfall characteristics is rainfall

intensity, usually expressed in millimeters per hour (mm/hr). Very
intense storms are not necessarily more frequent in areas having a
high total annual rainfall. Storms of high intensity generally last for
fairly short periods and cover small areas. Storms covering large
areas are seldom of high intensity but may last several days. The
infrequent combination of relatively high intensity and long duration
gives large total amounts of rainfall. These storms do much erosion
damage and may cause devastating floods. These unusually heavy
storms are generally associated with warm-front precipitation. They
are most apt to occur when the rate of frontal movement has
decreased, when other fronts may pass by at close intervals, when
stationary fronts persist in an area for a considerable period, or when
tropical cyclones move into the area.

79
1200

1000 1970

800 1960
Slope
600 1950

400 1945

200 1940

0
200 400 600 800 1000 1200

Accumulated Precipitation (cm)

 Intense rainstorms of varying duration occur from time to time over

almost all portions of the Philippine Archipelago, especially the
province of Northern Samar; however, the probability of these heavy
rainfalls varies with the locality. The first step in designing a water-
control or impounding facility is to determine the probable
recurrence of storms of different intensity and duration so that an
economically sized structure can be provided. For most purposes, it
is not feasible to provide a structure that will withstand the greatest
rainfall that has ever occurred. It is often more economical to have a
periodic failure than to design for a very intense storm. Where human
life is endangered, however, the design should handle runoff from
storms even greater than have been recorded. For these purposes, data
providing return periods of storms of various intensities and
durations are essential. This return period, also called recurrence
interval, is defined as the period within which the depth of rainfall for a
given duration will be equaled or exceeded once on the average.

 A general expression for rainfall intensity as a function of rainfall

duration and return period is given by

80
 i = K (Tx/tn)
(3.3)

where i is the rainfall intensity (mm/hr) , t is the rainfall duration

(min), T is the return period (years), and K, x, & n are constants for
a given geographical location.

In the Philippines, the Hydrology and Flood Forecast Center of

the Philippine Atmospheric, Geophysical and Astronomical Services
Administration (PAGASA) utilized the “Analysis of Extreme Value”
procedure developed by E. J. Gumbel from a study of the frequency
of large floods to estimate the rainfall intensity – duration –
frequency data of the country. Tables 3.1 and 3.2 present the rainfall
intensity – duration – frequency data for the University of Eastern
Philippines (UEP), Catarman, Northern Samar, based on 12 years of
records.

SAMPLE PROBLEMS

1. From the rainfall records of a PAGASA Station, you are given the following data:

Intensity (I, mm/hr) Duration (t, hrs) Frequency (T, yrs)

150 2.0 10
200 3.5 25
300 4.0 50

Assume that the equation of the form, I = k Tx / tn is applicable. Determine the values of “x”, “n” and “k”.
Using the derived equation, answer the following problems :
a. What is the rainfall intensity (mm/hr) for a 4-hr rainfall which recurs every 10 years ?
b. What is the duration (hrs) of a 200 mm/hr rainfall event that will recur every 20 years ?
c. What is the frequency (yrs) of a 2-hr rainfall event with rainfall depth of 200 mm ?
d. What is the rainfall intensity (in/hr) of a 3-hr rainfall which recurs every 25 years ?
e. What is the rainfall depth (in) of a 5-hr rainfall event which happens every after 50 years ?

Given Required

As stated in the problem. ● Derivation of the equation

● Answer the following problems:
a. I (mm/hr) when t = 4 hrs, T = 10 yrs
b. t (hrs) when I = 200 mm/hr, T = 20 yrs
c. T (yrs) when t = 2 hrs, d = 200 mm
d. I (in/hr) when t = 3 hrs, T = 25 yrs
e. d (in) when t = 5 hrs, T = 50 yrs

Solution

Assuming that the equation given in the problem is applicable, this is linearized through
logarithmic transformation:

81
 I = k Tx / tn

ln I = ln k + ln Tx – ln tn

Substituting the given values of I, T and t in the linearized equation, then –

ln 150 = ln k + x ln 10 – n ln 2
ln k + 2.303 x – 0.693 n = 5.011
(1)

ln 200 = ln k + x ln 25 – n ln 3.5
ln k + 3.219 x – 1.253 n = 5.298
(2)\

ln 300 = ln k + x ln 50 – n ln 4
ln k + 3.912 x – 1.386 n = 5.704
(3)

The equations are solved simultaneously by elimination to solve for the values of “k”, “x” and “n”. First
of all, eliminate “ln k” by subtracting equation (1) from equation (2), thus:

ln k + 3.219 x – 1.253 n = 5.298

− ln k + 2.303 x – 0.693 n = 5.011
0.916 x – 0.560 n = 0.287
x – 0.611 n = 0.313
(4)

Again, eliminating “ln k” by subtracting equation (2) from equation (3):

ln k + 3.912 x – 1.386 n = 5.704

− ln k + 3.219 x – 1.253 n = 5.298
0.693 x – 0.133 n = 0.406
x – 0.192 n = 0.586
(5)

From equation (4), “x = 0.611 n + 0.313” and substituting this in equation (5):

0.611 n + 0.313 – 0.192 n = 0.586

0.611 n – 0.192 n = 0.562 – 0.313
0.419 n = 0.249
n = 0.594

To solve for x substitute n = 0.594 in eq 5

x – 0.192 n = 0.586
x - 0.192 (0.594) = 0.586
x = 0.70

To solve for k substitute the value of n and x in eq. 3

ln k + 3.912 x – 1.386 n = 5.704

exponentiating:

e
lnk
= e 3.7878

k = 44.0975

82
a. t = 4 hours = 240 min. T = 10 years, what is I

Substitute in equation:

I = k Tx / tn

I = 44.0975(10)0.70/ 2400.594 ¿ ¿

I = 221.011/25.93

I = 8.5233 mm/hr.

b. t (hrs) when I = 200 mm/hr and T = 20 years

I = k Tx / tn

t = k Tx/ I
n

t
0.594
= 44.0975 ( 20)0.70/ 200

t
0.594
= 1.795

t = 2.6774 hrs.

Note: You may continue solving the required solution for letters c, d, and e using the derived
equation.

3.6.3 HYDROLOGIC FREQUENCY ANALYSIS

 It has been recognized that hydrologic events such as rainfall and

streamflow exhibit a high degree of variability. This variability has,
in the past, led to the extensive study of hydrologic processes.

 Hydrologic processes are generally analyzed probabilistically. This

deals with the estimate of the chance or likelihood of occurrence of a
given event by determining the frequency curve of best fit to samples
of hydrologic data, such as rainfall. A frequency curve relates the
magnitude of a variable to its frequency of occurrence. The curve is
an estimate of the cumulative distribution of the population of that
variable and is prepared from a sample of data. Frequency curves
are widely used in predicting droughts, storages, rainfall, wave
height and water quality. Knowledge of the frequency of hydrologic
events is necessary for the design of structures such as flood control
reservoirs, open ditches, farm ponds, tile drainage systems, bridge
openings, and many other water carrying structures.

 There are a number of frequency distributions used in hydrology.

Among these are the normal, log-normal, gamma, Poisson and other

83
 Pearson types of distribution functions. For this course, only the
normal distribution will be discussed because many phenomenon in
nature follow this distribution and hydrologic events are not
exceptions. The variability of the normal distribution is indicated by
the slope of the cumulative distribution; that is, the greater the
variability, the greater the slope. The standard deviation is equal to
the difference between magnitudes at probabilities of 16% and 50%
or 50% and 84%. The mean of the distribution occur at 50%
probability.

 The procedure involved here is called hydrologic frequency

analysis. By this method, the rainfall data are treated as statistical
variables. The frequency distribution of the data is examined and the
magnitude of the variable of a given recurrence interval or
probability of occurrence is determined. Recall that the recurrence
interval (T) is defined as the average interval of time within which
the magnitude of a given variable is equaled or exceeded.
Conversely, the probability of occurrence (P) is defined as the
percentage of observed events that are equal to or greater than, a
given event within a period of time. Recurrence interval and
probability of occurrence are inversely proportional as expressed in
the simple equation

P = 1/T (3.4)

The steps involved in hydrologic frequency analysis may be

outlined as:

1. Selection of data. There are two methods of data selection, the

annual series and the partial duration series. In the annual series,
only the largest event for each year is selected for analysis.
When design is controlled by the most critical conditions, such as
the design of spillway, the annual series should be used. With the
partial duration series, all values above a given base are chosen
regardless of the number within a given time period. The partial
duration series should be used if the second largest value (or
lower) of the year could affect the design of structures. An
example is the design of farm drainage in which the damage may
be due to the extent of flooding, which in turn may be caused
largely by associated peak flows. The annual and partial duration

84
 series of data selection give essentially identical results for
recurrence intervals greater than ten (10) years.

2. Determination of statistical parameters. Following the

selection of data, the next step is to determine values of the
statistical parameters for the analysis. The different methods of
analysis will involve the following:

a. Rank the events from the highest to the lowest with the largest
event being given a rank, m = 1; the second largest event, m = 2;
etc.

b. Compute the arithmetic mean (M)

M = X / N
(3.5)
where X = magnitude of variable or event and N = sample
size.

c. Compute the standard deviation (S)

 X2 – (X)2 / N
S=
(3.6)
N–1

d. Use an arithmetic-probability plotting paper to plot the mean (M)

at 50% probability (P = 0.5), the value of M + S at 15.9%
probability (P = 0.159) and the value of M – S at 84.1%
probability (P = 0.841). These three points should form a straight
line.
3. Fitting of data. To get a rough estimate of the goodness of fit of the
distribution to the data, plot the probability of occurrence (P) of the
events using the equation:

P = m / (N + 1)
(3.7)

The goodness of fit in this case is usually an eyeball estimate.

Where the fit is satisfactory, the resulting straight line fits the data
and this may be used for predicting the frequency of occurrence of
an event of a given magnitude.
SAMPLE PROBLEM

85
1. Perform a hydraulic frequency analysis on the data given. Assume that the frequency
distribution is normal.

Year Annual Year Annual Year Annual

rainfall Rainfall Rainfall
1980 2750 1988 2695 1996 3740
1981 3575 1989 3795 1997 2350
1982 3590 1990 3600 1998 3590
1983 2600 1991 3950 1999 2920
1984 2985 1992 3025 2000 2026
1985 3340 1993 2275 2001 2763
1986 2005 1994 2580 2002 2954
1987 2775 1995 2960 2003 3254

Determine the following:

a. The probability occurrence of a 3200mm rainfall.

b. The probability of the annual rainfall that will occur every 12 years.
c. The frequency of an annual rainfall of 2800 mm rainfall.
d. The magnitude of an annual rainfall having 50% probability of occurrence.
e. The frequency of annual rainfall of 2780 mm.

Solution:

Annual Rainfall

(x) (x2)

2750 7562500
3575 12780625
3590 12888100
2600 6760000
2985 8910225
3340 11155600
2005 4020025
2775 7700625
2695 7263025
3795 14402025
3600 12960000
3950 15602500
3025 9150625
2275 5175625
2580 6656400
2960 8761600
3740 13987600
2350 5522500
3590 12888100
2920 8526400
2026 4104676
2763 7634169
2954 8726116

86
 3254 10598516
_____ __________
72,097 223,727,577

M=
∑x =
72097
= 3004.04
n 24

S = √ ∑ x 2−¿ ¿ ¿ ¿ ¿

√
2
(72097)
S = 223727577− 24
24−1

S = 557.37

M+ S = 3004.04 + 557.37

= 3561.41 @ 15.9%

M – S = 3004.04 – 557.37

= 2446.67 @ 84.1%

Based on the graph ( See separate graph)

Probability of occurrence of 3200 is 40%

1 1
b. P = = (100)
T 12

P = 8.33%

1
c. P = (100)
T

100
64% =
T

T = 1.39 years

d. From the graph (see separate graph)

87
e. From the graph (see separate graph)

1
P= (100)
T

100
65% =
T

T = 1.59 years

3.6.4 POINT RAINFALL ANALYSIS

 A typical recording rain gage chart should be properly analyzed in

order to obtain needed data on rainfall depth or rainfall intensity
during a particular rainfall event. The line on the chart is a
cumulative rainfall curve, the slope of which being proportional to
the intensity of the rainfall. The peak is the point of reversal of the
recording gage. To analyze the chart, the time and amount of rain
should be selected from representative points where the rainfall rate
changes so that the data will represent the curve on the chart.

 For example, for the given rain gage chart, representative points
were selected. The cumulative rainfall and time for these points
were obtained and tabulated (Table 3.3). The time interval between
two representative points and the rainfall during the time interval
were determined and used to compute the rainfall intensity for said
interval.

 From the rain gage chart analysis, the maximum rainfall intensity at
different time durations can be calculated by the weighted average.
Mass rainfall curves, required for some types of analyses, may be
obtained by plotting the cumulative rainfall against cumulative time
while rainfall intensity histograms can be drawn by plotting a bar
graph of rainfall intensity versus its time increment (Fig. 3.5).

Table 3.3. Rain gage chart analysis.

Time Cumulative Cumulative Rainfall Time Rainfall Intensity

(AM) Rainfall Time During Interval for Interval
(mm) (min) Interval (mm) (min) (mm/hr)
6:50 0 0
1 10 6
7:00 1 10
10 10 60
7:10 11 20

88
 11 5 132
7:15 22 25
46 20 138
7:35 68 45
19 10 114
7:45 87 55
31 40 47
8:25 118 95
6 45 8
9:10 124 140
6 100 4
10:50 130 240

SAMPLE PROBLEM

1. Make a point rainfall analysis on the data below

Time (am) Cumulative Rainfall Time (am) Cumulative rainfall

4:05 11.00 7:15 268.32
4:15 27.75 7:26 298.15
4:25 69.30 7:43 319.50
4:45 83.70 8:05 335.90
5:15 90.00 8:15 360.65
5:25 123.75 8:35 385.45
5;40 196.25 8:45 393.60
6:02 215.00 9:05 405.10
6:17 223.75 9:25 415.20
7:00 246.02 9:45 419.60

Compute the following:

1. Maximum rainfall intensity (mm/hr.) for a duration of 45 min.
2. Maximum rainfall intensity (cm/hr) for a duration of 35 min.
3. Maximum rainfall depth (cm) for a one hour rain duration
4. Maximum rainfall depth for 1.5 hr. duration

Solution:

Time (am) Time Interval Cumulative mm/hr.

Min. Rainfall(mm)

4:05 11.00
10 16.75 100.50
4:15 27.75
10 41.55 249.30
4:25 69.30
20 14.40 43.20
4:45 83.70

89
 30 6.30 12.60
5:15 90.00
10 33.75 202.50
5:25 123.75
15 72.50 290.00
5:40 196.25
22 18.75 51.13
6:02 215.00
15 8.75 35.00
6:17 223.75
43 22.27 31.07
7:00 246.02
15 22.30 89.20
7:15 268.32
11 29.83 162.70
7:26 298.15
17 21.35 75.35
7:43 319.50
22 16.40 44.72
8:05 335.90
10 24.75 148.50
8:15 360.65
20 24.80 74.40
8:35 385.45
10 8.15 48.90
8:45 393.60
20 11.50 34.50
9:05 405.10
20 10.10 30.30
9:25 415.20
20 4.40 13.20
9:45 419.60

15 (290 )+ 10 ( 202.50 ) +20(51.13)

1. i45 =
45

= 164.40 mm/hr

15 (290 )+ 10 ( 202.50 ) +10(51.13)

2. i35 =
35

= 196.75 mm/hr. ( 1cm/10mm)

= 19.67 cm/hr.

3. d60 = 33.75 + 72.5 + 18.75 + 8.75

= 133.75 mm ( 1cm/10mm)

90
 = 13.375 cm

4. d90 = 72.5 + 18.75 + 8.75 + 22.27

= 122.27 mm

3.6.5 CLASSIFICATION OF RAINFALL EVENTS

 Since no two rainfall events have exactly the same time-intensity

relationships, it is often convenient to group storms with regard to
their characteristics. The most common characteristics used in such
groupings are the intensity of the rain and the pattern of the rainfall
intensity histogram.

 The pattern of a storm is determined by the arrangement of the

rainfall intensity histogram. Storm patterns are important because
they are one of the factors determining the shape of the runoff
hydrograph. The different storm patterns are shown in Fig. 3.6. The
advanced pattern of rainfall brings higher intensities when the
infiltration rate is the greatest, thus causing some reduction in the
runoff peaks. On the other hand, the delayed pattern causes higher
runoff peaks, as the high intensities occur when the infiltration is at a
minimum and depressional storage has been largely satisfied. In
general, the cold front produces a storm of an advanced type, and the
warm front a uniform or intermediate pattern.

3.6.6 ESTIMATING AVERAGE RAINFALL DEPTH OVER AREA

 Most hydrologic and agronomic problems require the estimation of

the average rainfall depth over an area such as a watershed and a
cropped area. The usual procedure for doing this is to extrapolate
gage measurements to areal averages. The most common methods
for accomplishing this include:

1. Arithmetic Averages. The simplest of the area averaging

methods, this involves averaging arithmetically all the rain
amounts measured by the rain gages within the area. This
method, however, assumes uniform gage coverage of the area and
linear variations in rainfall in between any two gaging stations.

2. Thiessen Method. The method assumes that the rainfall in an

area may be taken as similar to that recorded in the nearest gaging
station. The location of the rain gages in and around the area of
91
 interest are first plotted and straight lines connecting neighboring
stations are then drawn. On each of these lines, perpendicular
bisectors are drawn resulting in the formations of polygon
referred to as Thiessen polygons. Any area within a polygon is
physically nearest to the gaging station enclosed by that polygon
and, hence, assumed to have rainfall characteristics similar to that
recorded in that particular rain gage. The average rainfall over
the area is then estimated as the area weighted average rainfall of
all the polygons.

3. Isohyetal Method. In this method, the rainfall depths recorded in

all the stations in and around an area of interest are plotted on a
map of desirable scale. By interpolation, lines of equal rainfall
depths or isohyets are then plotted. In plotting the isohyets,
considerations may be given to various factors influencing the
precipitation over an area (e.g., orographic effects). The rainfall
within an area enclosed by two adjacent isohyets is taken as the
average value of the two isohyets. Hence, the average rainfall
over an area is the area weighted average rainfall of all the area
enclosed by two adjacent isohyets.

 In Table 3.4, the advantages and disadvantages associated with each of

the above-mentioned methods of estimating the average precipitation
over an area are tabulated.

Table 3.4. Comparison of the commonly used methods for

determining the average rainfall over an area.

Method of ADVANTAGES DISADVANTAGES

Determining
Average Rainfall
over Area
Arithmetic Average 1. Fast and easy. 1. Ignores orographic
2. Objectives effects.
3. Accurate to extent 2. Uses only stations in the
controlled by gage basin.
density. 3. Gives poor results if
4. Subject to only slight error stations unequally
in computation. distributed (which is the
general case).

Thiessen Method 1. Stations weighted 1. Slow.

according to area 2. Orography generally
controlled. ignored.
2. Utilizes stations outside 3. Stations have fixed areas

92
 basin. of control.
3. Objective. 4. Requires more
4. Requires some skills at computations. Therefore
drawing Thiessen subject to more errors.
polygons.

Isohyetal Method 1. Possible to account for 1. Subjective.

topography. 2. Slow.
2. Gives best results (most 3. Requires some training
accurate). and skill.
3. Takes advantage of 4. Subject to possible large
unequal gage density. errors due to number of
4. Utilizes stations out-side computations.
basin.
5. For large basins, it is
possible to determine
contributing area. Very
important in streamflow
analysis.
6. Results can be utilized in
transportation techniques.
7. Possible at a glance to see
rainfall pattern.

LEARNING ACTIVITIES :

I. Answers the following questions below:

1. Liquid water droplets usually It consists of tiny with diameters between 0.1 to 0.5 mm.
2. A type of precipitation typical of the tropics.
3. This type of precipitation results from the moving of moist air masses over mountain barriers
4. This type of precipitation is usually highly localized and short duration.
5. A type precipitation influenced by topography.
6. This type of precipitation results from filling of air converging into a low area.
7. In most parts of the Philippines, more than 80% of the rainfalls are due to what?
8. It is characterized by widespread cloudiness, precipitation and moderate to strong surface
winds.
9. The influence of Intertropical Convergence Zone on Philippines weather is felt mainly during
what period ?
10. If the vertical depth of water is less than 0.05 mm, what is the term of the rainfall?
11. The period of time between two successive rainfall events is termed as what?
12. A method of averaging rainfall depth over an area which assumes uniform distribution of
gages over the area is the termed as what?
13. It is defined as the period within which the depth of rainfall for a given duration will be
equaled exceeded once on the average.
14. This is the procedure to be followed in adjusting precipitation data.
15. A line connecting points of equal rainfall depths is termed what?

93
II. PROBLEM SOLVING : Solve the following as directed

1. You are given the annual rainfall in Catubig, Northern Samar, as follows :

Year Annual Rainfall, mm Year Annual Rainfall, mm

1981 2008 1996 2468
1982 2112 1997 2776
1983 2345 1998 2890
1984 1975 1999 2980
1985 1864 2000 2788
1986 1684 2001 2010
1987 1846 2002 1911
1988 1468 2003 1191
1989 1566 2004 1786
1990 1995 2005 2121
1991 1897 2006 2223
1992 1651 2007 2552
1993 1775 2008 1808
1994 1339 2009 1777
1995 1448 2010 2023

Using the frequency curve, find the following :

a. The frequency of an annual rainfall with a magnitude of 990 mm;

b. The magnitude of an annual rainfall with a recurrence interval of 5 years;
c. The probability of occurrence of an annual rainfall of 2,500 mm;
d. The magnitude of an annual rainfall that will have a 12.5% probability of occurrence.

1. The following table presents the rainfall data for a rain gage chart analysis.

Time Cumulative Cumulative Rainfall Time Rainfall Intensity

(AM) Rainfall Time During Interval for Interval
(mm) (min) Interval (mm) (min) (mm/hr)
6:50 0 0

7:00 5 10

7:10 17 20

7:15 29 25

7:35 87 45

7:45 109 55

8:25 118 66

9:10 124 85

10:50 130 88

94
 Using the computed rainfall intensities, estimate the following :

a. The maximum 15 minute rainfall intensity;

b. The maximum 30 minute rainfall intensity;
c. The maximum 45 minute rainfall depth;
d. The maximum 60 minute rainfall depth.
e.

3. A map of a catchment area with the locations of the rain gaging stations is given in Figure 1.
The following information are given:

Station No. Rainfall (mm)

1 112
2 126
3 128
4 136
5 142
6 114
7 97
8 145

Compute the average aerial precipitation using the arithmetic method.

4. Construct the Thiessen polygons for the catchment area (shown in Figure 1) and compute the
average aerial precipitation.

5. Construct the isohyets for the area (use Figure 2) by using 10 mm isohyetal interval. Start
with 110 mm isohyet. Compute the average aerial precipitation using the following isohyetal
areas.

95
96