19‐Feb‐16

Analysis of Precipitation 
Data

Analysis of Temporal Distribution of

Rainstorm Event
- Only feasible for data obtained from recording gauges.

- Rainfall Mass Curve : A plot showing the cumulative rainfall

Cumulative Depth

depth over the storm duration

- Rainfall Hyetogragh : A plot of rainfall depth or Time

intensity with respect to time

Depth or
Intensity

- Instantaneous Rainfall Intensity,

(slope of the mass curve) Time

dP(t)
i(t) 
dt
P P (t  t )  P (t )
- Average Intensity in (t, t + t) is it  
t t

1
 19‐Feb‐16

Rainfall Mass
Curve &
Hyetograph

Autographic Chart

2
 19‐Feb‐16

Clock-Time vs. Rolling-Time Max Rainfall

Example (GEO Raingage N17 on 5 November 1993)

Time 15-min 5-min

Rainfall (mm) Rainfall (mm)
3:45
3:50 9.0
3:55 12.5
4:00 35.0 13.5
4:05 17.0 45.0
4:10 14.5
4:15 37.5 6.0
4:20 5.0
4:25 5.0
4:30 14.5 4.5

 Clock-time 15-min maximum rainfall depth = 37.5 mm

 Rolling-time 15-min maximum rainfall depth = 45.0 mm

Point Data Analysis
 Point precipitation data refers to precipitation
of a station. This data record could be in the
form of hourly, daily, monthly or annual
precipitation.
 Depending upon the nature of catchment and
its area, there could be as many gauging
stations as feasible.
 Before using rainfall data, it is necessary to
check the data for continuity & consistency
 Missing data
 Record errors

3
 19‐Feb‐16

Estimation of Missing Precipitation Record
 Some precipitation stations may have short
breaks in the records because of absence of
the observer or because of instrumental
failures.
 In the procedure used by the U. S. Weather
Bureau, the missing precipitation of a
station is estimated from the
observations of precipitations at some
other stations as close to and as evenly
spaced around the station with the
missing record as possible.

Estimation of Missing Precipitation Record
 Methods for estimating missing data
1. Arithmetic Mean Method
2. Normal Ratio Method
 The station whose data is missing is called
interpolation station and gauging stations whose
data are used to calculate the missing station data are
called index stations.
 If the normal annual precipitation (mean of 30
years of annual precipitation) of the index stations
lies within ±10% of normal annual precipitation of
interpolation station then we apply arithmetic mean
method to determine the missing precipitation record
otherwise the normal ratio method is used for this
purpose.

4
 19‐Feb‐16

Estimation of Missing Precipitation Record
Consider that record is missing from a station ‘X’. 
Now let,
N = Normal annual precipitation
P = Storm Precipitation
Let Px be the missing precipitation for station ‘X’
and Nx , the normal annual precipitation of this
station, Na , Nb , and Nc are normal annual
precipitation of nearby three stations, A, B and C
respectively while Pa , Pb and Pc are the storm
precipitation of that period for these stations.

Estimation of Missing Precipitation Record
Now we have to compare Nx with Na , Nb and
Nc separately. If difference of Nx ‐ Na , Nx –
Nb , and Nx – Nc is within ±10% of Nx , then
we use simple arithmetic means method
otherwise the normal ratio method is used.

5
 19‐Feb‐16

Simple Arithmetic Mean Method
 Select rain gauge index stations (e.g. 1, 2,
& 3) as close to and as evenly spaced
around the station with the missing
rainfall record (i.e. station X)
 Collection of rainfall data for these
stations on the day for which data at
station X is missing
 Normal annual rainfall at all these
stations should also be collected.

Simple Arithmetic Mean Method
According to the arithmetic mean method the
missing precipitation ‘Px’ is given as:

In case of three stations 1, 2 and 3,

P1  P2  P3
Px 
3

6
 19‐Feb‐16

Normal Ratio Method
 If normal annual rainfall at one or all of these
three index stations differs more than 10% of
normal annual rainfall of station X, then
normal ratio method is applicable.
1 N N N 
Px   P1 x  P2 x  P3 x 
3  N1 N2 N3 

 Where P1, P2, P3 & Px represent respective

rainfall data at stations 1, 2, 3, and X respectively
of the day for which data is missing at station X,
and N1, N2, N3 & Nx represent their respective
normal annual rainfalls.

Example
Example:
Find out the missing storm precipitation of station ‘C’ given
in the following table:
Station A B C D E
P, Storm precipitation (cm)  9.7 8.3 ‐‐‐ 11.7 8.0
N, Normal Annual precipitation (cm) 100.3 109.5 93.5 125.7 117.5

Solution:
10% of Nc = 93.5 x 10/100 = 9.35
±10% Range of Nc= 84.15 to 102.85
Values of Nb , Nd and Ne are out of this range, hence normal
mean method applicable

7
 19‐Feb‐16

1 N N N N 
Pc   Pa c  Pb c  Pd c  Pe c 
4  Na Nb Nd Ne 

Pc = (1/4) (9.7 x 93.5/100.3 + 8.3 x 93.5/109.5 + 
11.7 x 93.5/125.7 + 8.0 x 93.5/117.5) = 7.8 cm.  

Example
Example:
Precipitation station “X” was inoperative for part of a
month during which a storm occurred. The storm
totals at three surrounding stations A, B and C were
respectively 10.7, 8.9 and 12.2 cm. The normal annual
precipitation amounts at stations X, A, B and C are
respectively 97.8, 112, 93.5 and 119.9 cm. Estimate the
storm precipitation for station “X”.

8
 19‐Feb‐16

Example
Solution:
Pa = 10.7 cm Na = 112 cm
Pb = 8.90 cm Nb = 93.5 cm
Pc = 12.2 cm Nc = 119.9 cm
Px = ? Nx = 97.8 cm
10% of Nx = 97.8 x 10/100 = 9.78 cm
Nx – Na = 97.8 – 112 = ‐14.2 cm (more than +10% of Nx,
therefore no need to calculate Nx – Nb and Nx – Nc )
So we will use Normal Ratio Method.
Px = (1/3) (10.7 x 97.8/112 + 8.90 x 97.8/93.5 + 12.2 x 
97.8/119.9) = 9.5 cm.

Consistency of Precipitation Data or 
Double Mass Analysis
Some of the common causes for inconsistency of
record are:
(i) shifting of a rain gauge station to a new
location,
(ii) the neighborhoods of the station undergoing
a marked change,
(iii) change in the ecosystem due to calamities,
such as forest fires, land slides,
(iv) occurrence of observational error from a
certain date,
(v) Any other
18

9
 19‐Feb‐16

Test for Consistency of Record 

Accumulated Annual Rainfall at x  ΣP in 
units of 103 cm 

Mc
Pcx  Px
Ma

Accumulated Annual Rainfall of 10 station Mean
ΣP in units of l03 cm 
19

Double‐mass curve technique
 The checking for inconsistency of a
record is done by the double‐mass
curve technique. This technique is
based on the principle that when each
recorded data comes from the same
parent population, they are consistent.

10
 19‐Feb‐16

Test for consistency record
(Double mass curve techniques)

• Let a group of 5 to 10 base stations in the neighborhood of the

problem station X is selected
• Arrange the data of X station rainfall and the average of the
neighboring stations in reverse chronological order (from
recent to old record)
• Accumulate the precipitation of station X  Px  and the
average values of the group base stations  Pavg  starting from
the latest record.
• Plot the  Px  against  Pavg  as shown on the next figure
• A decided break in the slope of the resulting plot is observed
that indicates a change in precipitation regime of station X, i.e
inconsistency.
• Therefore, it should be corrected by the factor shown on the
next slide

Double Mass Curve Analysis Test for consistency record….

5
accumulated annual rainfall of X stn in 10^3 cm

4.5

3.5
Mc c
3
c
a

2.5 Ma a
2

1.5
Mc
1 Pcx  Px
0.5 Ma
0
0 0.5 1 1.5 2 2.5
Accumulated annual rainfall of neigbouring stns in 10^3 cm

Pcx – corrected precipitation at any time period t1 at station X

Px – original recorded precipitation at time period t1 at station X
Mc – corrected slope of the double mass curve
Ma – original slope of the mass curve

11
 19‐Feb‐16

Example
Example:
Check consistency of the data given in
table below and adjust it if it is found
to be inconsistent.

Table Precipitation Data

Year Annual Mean of annual Year Annual Mean of annual
precipitation at x precipitation of 20 precipitation at x precipitation of
(mm) surrounding stations (mm) 20 surrounding
(mm) stations (mm)

1972 188 264 1954 223 360

1971 185 228 1953 173 234
1970 310 386 1952 282 333
1969 295 297 1951 218 236
1968 208 284 1950 246 251
1967 287 350 1949 284 284
1966 183 236 1948 493 361
1965 304 371 1947 320 282
1964 228 234 1946 274 252
1963 216 290 1945 322 274
1962 224 282 1944 437 302
1961 203 246 1943 389 350
1960 284 264 1942 305 228
1959 295 332 1941 320 312
1958 206 231 1940 328 284
1957 269 234 1939 308 315
1956 241 231 1938 302 280
1955 284 312 1937 414 343

12
 19‐Feb‐16

Solution:
A double mass curve is plotted by taking cumulative of average
precipitation of surrounding stations along x-axis and
accumulative precipitation of station ‘X’ along
y-axis for which consistency of data is being investigated. The
double mass curve is shown in Figure below:

12000

10000
8000 Sa/S0=0.7
6000
S0=1.176
4000

2000 Sa=0.854

0
0 2000 4000 6000 8000 10000 12000

Double Mass Curve

Slope of Ist line = Sa = 0.854

Slope of deviating line = So = 1.176
Correction to values (multiplying factor) = 
0.854/1.176 = 0.70

13
 19‐Feb‐16

Table Adjusted Precipitation
Year Cumulative Annual precipitation Cumulative precipitation of 20 Corrected Precipitation Remarks
at x(mm) surrounding stations (mm)

1972 188 264 188

1971 373 492 185
1970 683 878 310
1969 978 1175 295
1968 1186 1459 208
1967 1473 1809 287
1966 1656 2045 183
1965 1960 2416 304

No Correction
1964 2188 2650 228
1963 2404 2940 216
1962 2628 3222 224
1961 2831 3468 203
1960 3115 3732 284
1959 3410 4064 295
1958 3616 4295 206
1957 3885 4529 269
1956 4126 4760 241
1955 4410 5072 284
1954 4633 5432 223
1953 4806 5666 173
1952 5088 5999 282
1951 5306 6235 218
1950 5552 6486 246
1949 5836 6770 198.8

Precipitation of station of
1948 6329 7131 345.1
1947 6649 7413 224
1946 6923 7665 191.8
1945 7245 7939 225.4

"X" x 0.7
1944 7682 8241 305.9
1943 8071 8591 272.3
1942 8376 8819 213.5
1941 8696 9131 224
1940 9024 9415 229.6
1939 9332 9730 215.6
1938 9634 10010 211.4
1937 10048 10353 289.8

14