Precipitation

1
Precipitation is the release of water form the atmosphere to reach the surface of Earth. The planning
assessment and management of water resources depend on the knowledge of the form and amount
of precipitation occurring in a given region of consideration over a time period of interest. Rivers of
north India are perennial, in peninsula region the river receives runoff due to rainfall, so they often go
dry in the summers

Forms of Precipitation:
• Rain: this includes drops of water falling from clouds having a diameter of at least 0.5 mm.
• Snow: These are precipitation in the form of ice crystals
• Hail: Precipitation in the form of rounded, or irregular pieces of ice over 5mm in diameter is
called hail. Occurs during violent thunderstorms
• Sleet: These are small, clear to translucent particles of ice and are generally 1mm-4mm in
diameter.
• Drizzle: These are fine, smaller droplets of rain having diameter range between 0.1 to 0.5 mm.
• Glaze: It is the ice coating formed when freezing raindrops turns to ice on coming in contact
with cold objects at the ground.
2
❑ Precipitation may be classified into following categories

Convective Precipitation Orographic Precipitation

• These types of precipitation is brought about by the • This type of precipitation results from the mechanical
heating of the air at the interface with the ground. lifting of the moist air over the mountain barriers and
causes heavy precipitation.
• This heated air expands because of its lesser density.
Air from cooler surroundings flow to take the • Factors important in the process are land elevation,
‘emptied’ place, hence forming a convective cell. slope, orientation of land slope, distance from
moisture source etc.
• The warm air continues to rise and dynamic cooling
takes place eventually causing condensation and • Such precipitation is witnessed in the lower Himalayas
precipitation. region and the Western Ghats of India.

Figure 1. Convective precipitation Figure 2. Orographic precipitation

3
❑ Rain shadow effect: Where there is a large and high land mass it is common to find the rainfall considerably higher on one
side than the other. This is through a combination of altitude, slope, aspect and dynamic weather direction influences and
can occur at many different scales.
Case Study: South Island of New Zealand
• The predominant weather pattern for the South Island of New Zealand is a series of rain-
bearing depressions sweeping up from the Southern Ocean, interrupted by drier blocking
anticyclones.

• The Southern Alps form a major barrier to the fast-moving depressions and have a huge
influence on the rainfall distribution within the South Island. Alps stretch the full length of
the South Island (approximately 700 km) and at their highest point are over 3,000 m above
mean sea level.

• The rain shadow effect can be clearly seen with the west coast rainfall being at least four
times that of the east.

• Table 2.2 also illustrates the point, with the number of rain days at different sites in a cross
section across the South Island.

• This pattern of rain shadow is seen at many different locations around the globe.

4
Cyclonic Precipitation
The uneven heating of the Earth’s surface by the sun results in the
high-low pressure difference regions. Cyclonic precipitation is
associated with lifting of air mass from high pressure region to low
pressure region. These are of two types:

✓ Frontal Precipitation: It is caused by the movement of warm

air on one side over the cold air. If the masses are moving so
that warm air replaces the cold air, this front is known as
warm front. If the cold air displaces the warm air, it is said to
be cold front.

✓ Non-frontal Precipitation: It is a result of air flowing from

the surrounding high pressure region to low pressure region,
the lifted warm air cools down and precipitate.

This precipitation is also called hurricane or typhoon, accompanied

with high velocity wind and intense storm, having a diameter range
of 300-1500km. The extra-tropical cyclone covers large diameter, Figure 3. Cyclonic precipitation
which is up to 3000 km

5
❑ Measurement of Precipitation
• Precipitation is measured by the network of rain gauges and snow gauges.
While installing rain gauges, following points should be considered due to
errors
- The site of installation should be a flat and open ground. This site should
not be close to any forested area or building.
- The height of the fencing and should not exceed twice its distance from
the gauge.
- The installation of rain gauge should be avoided at the steep hill side.
• The four main sources of error in measuring rainfall that need consideration
in designing a method for the accurate measurement of rainfall are: Fig. Surface rain gauge with non-splash surround

- Evaporation losses: Losses due to evaporation

- Wetting loss: Losses due to wetting of the gauge
- Rain splash: Over-measurement due to splash from the surrounding
area
- Turbulence around a raised gauge: Under-measurement due to
turbulence around the gauge

Fig. Baffles surrounding a rain gauge to lessen the impact of

Fig. The effect of wind turbulence on a raised rain gauge. An area of reduced pressure (and uplift)
develops above the gauge in a similar manner to an aircraft wing. This reduces the rain gauge
wind turbulence.
6
catch.
 ❑ Measurement of Precipitation

Types of rain gauges:

Non-recording rain gauge : These types of rain gauges does not

record the data and collects only rainwater.
• Symon’s rain gauge is one of the expensively used rain gauges
in India.
• It consists of a collector of 12.7 cm diameter to a funnel. In
the metallic container, there is a funnel that discharges the
rainfall into vessel. Water collected in the bottle Is measured
using a suitably graduated measuring jar with 0.1 mm
accuracy.
• In India, rainfall is measured everyday at 8:00 AM IST and is
recorded as the rainfall of that day.
• The non-recording rain gauge gives the total depth of rainfall
during last 24 hours.
Figure 4. Schematic view of Symon’s rain gauge

7
Recording type Rain Gauge: These types of rain gauges has a mechanical arrangement by which the total
amount of rainfall since the time at which rainfall starts, gets automatically recorded on the graph paper with
time on the x-axis and amount of the rain is recorded on the y-axis. A mass curve is plotted, and various
information are derived.
These are of three types.

Tipping bucket type rain gauge:

• The rainwater from the funnel falls into one of a pair of small
buckets.
• Only one bucket will be I action at any given instant of time.
When one bucket receives 0.25 mm of rainfall, it tips and
brings the other bucket in position.
• The water from the container is collected in a tank below and
the process is repeated.
• This tipping of bucket actuates an electrically-driven pen to
trace a record on the chart mounted on a clock-driven drum.
• These types of rain gauges can be run using very small power
source and is very useful in hilly and inaccessible areas.

Figure 5. Tipping bucket type rain gauge

8
Weighing bucket type rain gauge: Float type ( natural syphon) rain gauge:
• In this type of rain gauges, the size of bucket is • This type of rain gauge is used by the IMD (Indian
designed to collect the maximum rainfall expected to Meteorological Department.
occur in a day. • In these types of rain gauges, the rain water is
• When a certain weight of water is collected in bucket, collected into float chamber via funnel.
the bucket acts as spring system, which makes the pen • In response to the amount of water collected, the
to move on a chart. float rise and a pen attached to the float through a
• The instrument gives a plot of cumulative rainfall lever system records the rainfall.
against time. • When the float fills up, a syphon arrangement
empties the float chamber.

9
Figure 6. Weighing bucket type rain gauge. Figure 7. Float type ( natural syphon) rain gauge:
Radar Measurement of Precipitation
• Radio Detective and Ranging (radar) can be used to measure precipitation over a large area-spatial extent.
• The electromagnetic energy released and received back by radar is measure of rainfall intensity.
• The major limitation of this technique is that images often do not reveal the rainfall-producing clouds because of
overlying cloud layers.

❑Rain Gauge Network :

• The number of rain gauges and their distribution affect the nature of collected rainfall data.
• The rain gauge density is defined as the ratio of the total area of catchment to the total number of rain-gauges in a
catchment. According to McCulloch (1961), the number of rain gauges (𝑁) are as following:
𝟐
𝑪𝒗
𝑵=
𝜺
Where 𝐶𝑣 is the coefficient of variation and 𝜀 is the percent allowable error, generally taken as 10%

10
• In this regard, the World Meteorological Organization has made the following recommendations for minimum
number of rain gauges in a catchment.
• Region I: Flat regions of temperate, Mediterranean and tropical zones: minimum one rain gauge station for 600-
900 𝑘𝑚2 is recommended. However, one rain-gauge station for 900-3000 𝑘𝑚2 is also acceptable.
• Region II: Mountainous regions of temperate, Mediterranean and tropical zones: Minimum one rain gauge station
for 100 to 250 𝑘𝑚2 is recommended. However, one rain-gauge station for 250-1000 𝑘𝑚2 is also acceptable.
• Region III: Arid and polar zones: Minimum one rain-gauge station for 1500 to 10,000 𝑘𝑚2 is acceptable.

Also, The Indian Standard Institute ISI (IS: 4987-1968) has recommended the following regarding minimum number of
rain-gauges.
• In plain regions, one rain-gauge per 520 𝑘𝑚2 of the catchment is recommended.
• In regions of average elevation, one rain-gauge per 260 to 390 𝑘𝑚2 of the catchment is recommended.
• In predominantly hilly areas with heavy rainfall. One rain gauge per 130 𝑘𝑚2 of the catchment is recommended.
In India, on an average, there is one rain gauge station for every 500 𝑘𝑚2 while in developed countries it is one rain
gauge station for 100 𝑘𝑚2 .

11
Example 2.1 In a catchment the annual rainfall recorded by six rain gauge stations are given in the Table below.
Determine the number of optimum stations in the catchment, for error of 10% in the estimation of the mean
rainfall.
Station A B C D E F
Rainfall 80 100.2 178.2 108.2 97.7 135.5

Solution :
Given, Number of stations, 𝑛 = 6; Allowable degree of error, 𝜀 = 10%

80+100.2+178.2+108.2+97.7+135.5
Mean precipitation, 𝑥ҧ = = 116.3 𝑐𝑚
6

80−116.63 2 + 100.2−116.63 2 + 178.2−116.63 2 + 108.2−116.63 2 + 97.7−116.63 2 + 135.5−116.63 2

Standard deviation, 𝜎 =
6−1

−36.63 2 + −16.43 2 + 61.57 2 + −8.43 2 + −18.93 2 + 18.87 2

= = 35.18
6−1

𝜎 35.18
Coefficient of variation 𝐶𝑣 = ∗ 100% = ∗ 100% = 30.16%
𝑥 116.63
𝐶𝑣 2 30.16 2
Optimum number of stations, 𝑁 = = = 9.1
≅ 10
𝜀 10
Hence, the optimum number of station for the given number of catchment is 10. therefore, four additional rain gauges are 12
needed.
❑ Rainfall Intensity
The intensity of rainfall is the rate at which rain occurs and is measured in the terms of depth per unit time. Based on
this, the rainfalls are classified into the following three categories.
• Light rain: Its falling rate is up to 2.5 mm/h.
• Moderate rain: Intensity ranges from 2.5-7.5 mm/h.
• Heavy Rain: Intensity is greater than 7.5 mm/h.

Estimation of missing data: Sometimes, due to instrument failure or absence of observer, it is possible that the
measurement of rainfall may consist of short breaks. Under such condition these methods are applied to calculate the
missing records.
• Arithmetic Average method: in this method, the value of missing rainfall point is calculated by taking average of the
station’s record. IF the missing value of rainfall at station X is 𝑃𝑥 and 𝑃1 , P2 , P3 … Pm are the rainfalls at the 𝑚
surrounding rain-gauges, then

𝟏
𝑷𝒙 = 𝑷 + 𝑷𝟐 + 𝑷𝟑 + ⋯ 𝑷𝒎
𝒎 𝟏
The rainfall records of at least three rain gauge stations should be there to yield good results. Limitation of this
method is that variation in normal annual precipitation at the adjacent stations and missing stations should be within
10%.
13
2. Normal ratio method: In this normal ratio method, the missing rainfall 𝑃𝑥 at station 𝑋 is estimated by weighing
normal annual precipitation of each index station.
Let, 𝑁1 , 𝑁2 , 𝑁3 … 𝑁𝑚 is the normal annual rainfalls at the 𝑚 surrounding rain gauges, respectively, 𝑁𝑥 be the normal
annual rainfall at station 𝑋 for which rainfall data 𝑃𝑥 is missed. Then,

𝟏 𝑵𝒙 𝑵𝒙 𝑵𝒙
𝑷𝒙 = 𝑷𝟏 + 𝑷𝟐 + … 𝑷
𝒎 𝑵𝟏 𝑵𝟐 𝑵𝒎 𝒎
This method is used when normal annual rainfall at surrounding stations differ from the annual rainfall station at
which rainfall data is missed by more than 10%.
3. Inverse distance square method: In this method, the missing rainfall is determined as a weighted average of the
observed rainfall at the surrounding stations. This method provides greater weights to the points closest to the
prediction location, and the weights diminish as a function of square of distance inverse.
Let, 𝑃1 , P2 , P3 … are rainfall at the stations 𝐴, 𝐵 and 𝐶 respectively, then the rainfall at station 𝑋 for which rainfall data
𝑃𝑥 is missed, is given by:

𝟏 𝟏 𝟏
𝑷 + 𝑷 + 𝑷
𝒓𝟐𝟏 𝟏 𝒓𝟐𝟐 𝟐 𝒓𝟐𝟑 𝟑
𝑷𝒙 =
𝟏 𝟏 𝟏
+ +
𝒓𝟐𝟏 𝒓𝟐𝟐 𝒓𝟐𝟑
Here,𝑟1 , 𝑟2 , and 𝑟3 are the distances from station 𝑋 whose data is to be estimated from data estimator station (𝐴, 𝐵
and 𝐶) . 14
4. Regression method: A multiple linear regression is assumed between rainfalls at the surrounding stations and
rainfall at station 𝑋 for which rainfall data 𝑃𝑥 is missed. This may be written as following

𝑷𝒙 = 𝒌𝟎 + 𝒌𝟏 𝑷𝟏 + 𝒌𝟐 𝑷𝟐 + ⋯ 𝒌𝒎 𝑷𝒎

The constants 𝑘0 , 𝑘1 … 𝑘𝑚 are obtained through regression analysis of previously recorded data.

Example 2.2 The normal annual rainfalls at the stations A,B, C and in a catchment are 780 mm, 830 mm,730 mm and
890 mm respectively. Out of these four rain gauge stations, it was found that rain-gauge station D did not function for a
part of month during a storm and the stations A, B, C recorded 90 mm, 80 mm and 100 mm rainfalls, respectively,
produced by the storm. Calculate the missing storm rainfall at station D.
Solution: 𝑁𝐴 = 780𝑚𝑚; 𝑁𝐵 = 830 𝑚𝑚; 𝑁𝑐 = 730 𝑚𝑚 and 𝑁𝐷 = 890 𝑚𝑚

𝑁𝐷 − 𝑁𝐴 890 − 780
× 100 = × 100 = 12.36%
𝑁𝐷 890

Similarly,
𝑁𝐷 − 𝑁𝐵 𝑁𝐷 − 𝑁𝑐
× 100 = 6.74% 𝑎𝑛𝑑 × 100 = 17.98%
𝑁𝐷 𝑁𝐷
Hence, the normal annual rainfall at the surrounding stations (A and C) differ from the annual rainfall of the station (D) at
which the rainfall data is missed by more than 10%. So, we need to use normal ratio method instead of arithmetic average
15
method.
Given,

Station Average annual rainfall(mm) Normal annual rainfall (mm)

A 90 780
B 80 830
C 100 730
D ? 890
Computation of missing data using normal ratio method:

1 𝑁𝐷 𝑁𝐷 𝑁𝐷
𝑃𝐷 = 𝑃𝐴 + 𝑃𝐵 + 𝑃
𝑚 𝑁𝐴 𝑁𝐵 𝑁𝐶 𝐶

890 90 80 70
𝑃𝐷 = + +
3 780 830 730

𝑷𝑫 = 𝟏𝟎𝟑. 𝟒𝟖 𝒎𝒎

Example 2.3. Consider station A to be a the origin of a rectangular coordinate system such that coordinates four
surrounding stations nearest to A in the quadrants are (20,30) km, (-10,15) km, (-15, -10)km and (10,-5)km,
respectively. Stations A failed to record the rainfall during the storm. Estimate the missing rainfall at A if the storm
rainfalls at the four surrounding stations are 75 mm, 90 mm, 70 mm and 60 mm, respectively.

16
Solution: Let 𝑟1 be the distance between station A and station in the first quadrant. Therefore,
𝑟12 = 202 + 302 = 1300
Similarly, for other stations,
𝑟22 = −10 2 +152 = 325
𝑟32 = −15 2 + −10 2 = 325
𝑟42 = 102 + −5 2 = 125
Using inverse least distance square method, missing rainfall of stations at station 𝐴 ,

1 1 1 1
𝑃 + 𝑃 + 𝑃 + 𝑃
𝑟12 1 𝑟22 2 𝑟32 3 𝑟42 4
𝑃𝐴 =
1 1 1 1
2+ 2+ 2+ 2
𝑟1 𝑟2 𝑟3 𝑟4

1 1 1 1
1300 75 + 90 + 70 + 60
𝑃𝐴 = 325 325 125
1 1 1 1
+ + +
1300 325 325 125

1.03
𝑃𝐴 = = 69.02 𝑚𝑚
0.014923

17
❑Mean Precipitation over an Area
• It is very important to know the areal distribution of precipitation for hydrologic analysis, water resources planning
or designing the collectors for a city drain system.
• The reliability of rainfall measured at one gauge in representing the average depth over a surrounding area is a
function of (a) the distance from the gauge to the center of representative area, (b) the size of the area, (c)
topography, (d) the nature of the rainfall of concern, etc.

✓ Arithmetic Average Method: The most direct approach is to use the simple arithmetic average of amounts of
rainfall at the individual rain gauge stations in the area. This procedure is satisfactory If gauges are uniformly
distributed and the topography is flat. If 𝑃1 , P2 , P3 …𝑃𝑚 are the rainfalls recorded at the stations 1,2,3 … 𝑚, then
the average depth of rainfall 𝑃ത is given as

𝟏
ഥ=
𝑷 (𝑷 + 𝐏𝟐 + 𝐏𝟑 …+𝑷𝒎 )
𝒎 𝟏
✓ Thiessen Polygon Method: This method attempts to allow for non-uniform distribution of gauges by providing
factor for each gauge. The area is divided into polygon subareas using the rain gauges at centers. The sub areas are
used as weights in estimating the watershed average depth.
In the figure, there are 7 rain gauge stations of which 4 are inside the boundary of catchment and 3 outside it. These
rain gauges are joined together forming a network of triangles. Perpendicular bisectors are drawn to each sides of all
the triangles and then joined together forming a polygon, one for each gauge. The area of polygon around each rain
gauge is measured.
18
 If 𝑃1 , P2 , P3 …𝑃𝑚 are the rainfalls recorded at the stations 1,2,3 … 𝑚, and 𝐴1 , 𝐴2 , 𝐴3 … 𝐴𝑚 are the area of
polygon enclosed by them, respectively then the average depth of rainfall 𝑃ത is given as

𝑷𝟏 𝑨𝟏 + 𝑷𝟐 𝑨𝟐 + 𝑷𝟑 𝑨𝟑 … . 𝑷𝒎 𝑨𝒎
ഥ=
𝑷
𝑨𝟏 + 𝑨𝟐 + 𝑨𝟑 … . +𝑨𝒎

This method has following advantages:

• It makes use of data from nearby stations located outside the catchment.
• It allocates importance of measurement according to the station spacing.

The limitations of this method are:

Figure 8. Thiessen Polygon Method:

• This method does not make any allowances for orographic influences in the basin. Thus it is not suitable to
compute the average areal rainfall of mountainous catchment.
• If a new rain gauge is added to the existing network or position of rain gauge is changed in the catchment, then
the network changes and new polygons are required to be sketched. Thus in this case, a fresh-computation of 19
✓ Isohyetal Method: This method used the technique of interpolation.
• The rain gauge locations are plotted on a suitable map and rainfall amounts are recorded.
• Then, interpolation between rain gauges is performed and rainfall amount at selected increments are
plotted.
• Identical depths from each interpolation are then connected to form isohyets (lines of equal rainfall depth.
• Once the isohyetal map is developed, the area 𝐴𝑗 between each pair of isohyets, within the catchment is
multiplied by the average 𝑃𝑗 of the rainfall depths of the two boundary isohyets to compute the areal
average precipitation as shown in the figure.

• This is the most flexible and accurate method for

determining the average precipitation over an area.

• However, it needs a fairly dense network of gauges to

correctly construct the isohyetal map and is time consuming
compared to other methods.

Figure 9.Isohyetal Method

20