Time Series Analysis of Rainfall in North Bangalore Metropolitan Region using Remote Sensing & Geographic Information System Techniques

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/280882089
Time Series Analysis of Rainfall in North Bangalore Metropolitan Region using
Remote Sensing & Geographic Information System Techniques
Article · March 2014
CITATIONS
0
READS
335
2 authors, including:
Some of the authors of this publication are also working on these related projects:
Aquifer performance test and Resistivity survey View project
T J Renuka Prasad
Bangalore University
16 PUBLICATIONS 12 CITATIONS
SEE PROFILE
All content following this page was uploaded by T J Renuka Prasad on 11 August 2015.
The user has requested enhancement of the downloaded file.

ISSN 2320-5407 International Journal of Advanced Research (2014), Volume 2, Issue 4 ,824-832
824
Journal homepage: http://www.journalijar.com INTERNATIONAL JOURNAL
OF ADVANCED RESEARCH
RESEARCH ARTICLE
Time Series Analysis of Rainfall in North Bangalore Metropolitan Region using Remote
Sensing & Geographic Information System Techniques
*Ramesh L Dikpal1
and T J Renuka Prasad2
1. Research Scholar, Department of Geology, Bangalore University, Bangalore, Karnataka, India
2. Professor, Department of Geology, Bangalore University, Bangalore, Karnataka, India.
Manuscript Info Abstract
Manuscript History:
Received: 12 February 2014
Final Accepted: 12 March 2014
Published Online: April 2014
Key words:
Bangalore Metropolitan
Region;Time series analysis,
ArcGIS, Auto correlation,
Geomorphology, Power Spectrum,
Moving Average
*Corresponding Author
Ramesh
Rainfall studies are of utmost utility for understanding nature & hence the
behaviour of climate changes. Time series is a set of observations taken at
specified times usually at equal interval. The inherent variability displayed
by many hydrological time series usually mask trends and periodic patterns.
This situation has often led to “something” the original time series so that the
effects of random variations are reduced and trends or cyclical patterns
enhanced. Thus a set of data depending on time is called a Time series. Here,
Rainfall series represent the time series. The time series analysis is helpful to
compare the actual performance and analyse the cause of variations. By
comparing different time series we can draw important conclusion. Graphical
method implies in increasing trend for pre-monsoon, south-west monsoon,
north-east monsoon and annually.Geo- informatics module consists of GIS
mapping for Location map, Geomorphology map and Season wise Rainfall
maps are generated. Autocorrelation indicates the periodicity observed as
37,16 & 6 years (PM), 12, 37 & 16 years (SWM), 8, 18 & 6 years (NEM)
and 16, 22 & 8 years (Annual) respectively. Power spectral depicts the
cyclicity of 37, 4 & 3 years (PM), 2, 4& 2 years (SWM), 3, 7 & 2 years
(NEM) and 2, 4 & 2 years (Annual) respectively. Moving average displays
prominent positive correlation coefficients at lags of 18 to 42 years in PM &
SWM and 12 to 24 years in NEM & Annual. The southwest and southeast
parts of the study area experience the heavy rainfall whereas the least rainfall
areas are the northern parts of the study area.The short term and long term
cyclicity observed in Autocorrelation, power spectrum and Moving Average.
Spatial variation of rainfall for the three seasons and annual has been studied
Copy Right, IJAR, 2014,. All rights reserved
Introduction
Changing precipitation pattern and its impact on surface water resources is an important climatic problem facing
society today. Associated with global warming, there are strong indications that rainfall changes are already taking
place on both the global and regional scales (Vyas et al., 2012). Spatial differences in trends can occur as a result of
spatial differences in the changes in rainfall and temperature and spatial differences in the catchment characteristics
that translate meteorological inputs into hydrological response (Burn and Elnur, 2002). Trend is present when a time
series exhibits steady upward growth or a downward decline, at least over successive time periods. Trend may be
loosely defined as “long-term change in the mean level”, but there is no fully satisfactory mathematical definition.
But trend analysis helps in finding „forecasting‟. The base of scientific forecasting is statistics. Trend analysis was
carried out to examine the long term trends in rainfall over different subdivisions. The rainfall trend is very crucial

825
for the economic development and hydrological planning for the country. Long term trends of Indian Monsoon
rainfall for the country as a whole as well as for smaller regions have been studied by several researchers.
In this paper, several approaches have been proposed for analysing time series such as Graphical, Auto Correlation,
Power Spectrum, Smoothed Power Spectrum and Moving Average method. Geo- informatics module consists of
GIS mapping for location map, Geomorphology map, Forest map, and Seasonal & Annual Rainfall maps are
generated.The purpose of this research is to detect best trend for the time series taken into account. In recent years
the techniques of forecasting have improved to a marked degree and are applicable everywhere. Though there are
several methods, techniques have been developed in finding the trend and forecasting, the finding suitable method is
an important task, because the rainfall trend is very crucial for the economic development and hydrological planning
for the country.
Material and Methods
The study area encompasses a geographical area of 4,665 km2
(Fig: 1) between 800 m and 1800 m altitude above
mean sea level and the study area lies between latitude 12.50N to 13.30N and longitude 77.00E to 78.10E in
Survey of India (1:50,000) Toposheet Nos. 57G/4, 57G/7, 57G/8, 57G/11, 57G/12, 57G/15, 57G/16, 57H/1, 57H/5,
57H/9 and 57H/13.
The study area covers seven taluks namely Bangalore North, Magadi, Nelamangala, Doddaballapura, Devanahalli,
Hoskote and Malur. For the present study monthly accumulated rainfall data was collected for the period of 1901 to
2010 from India Meteorological Department (IMD) and Karnataka State Natural Disaster Monitoring Centre
(KSNDMC), Bangalore and obtained seasonal and yearly rainfall over the region. The rainfall pattern in the study
area was classified as Pre-Monsoon (January-May), South-West Monsoon (June–September) and North-East
Monsoon (October–December).
The summary statistics like Mean, Median and Standard deviation have been analyzed to study the cyclicity of
rainfall to by Power spectrum; Moving Average and Autocorrelation methods for period of 110 years using
FORTRAN programme. The spatial variation of rainfall indifferent seasons and annual is obtained using GIS
platform. The forest region has been delineated using remotely sensed data. The geomorphology (Fig. 2) studies will
be correlated to rainfall for the water management purposes.
Result and Discussion
a. Geomorphology
Geomorphology reflects various landforms and structural features. Many of these features are favorable for the
occurrence of groundwater and are classified in terms of groundwater potentiality. These units are deciphered from
the remote sensing data, generated by using ArcGIS software are shown in Fig. 2. The major geomorphological
units found in the study area are Denudational Hills, Pediplain, Plateau and Structural Hills. These are briefly
described as follows:
Denudational hills (DH): These are formed due to differential erosion and weathering. Erosion and Weathering in
the study area are majorly depends on rainfall. Denudational hills occupy the northern and western edge of the area.
The groundwater prospect in his zone is negligible.
Pediplain: This unit will be developed as a result of continuous processes of pedimentation. The latitudinal
variations are relatively high for rolling plain to the extent of 5-10 m. These areas are described as nearly flat terrain
with gentle slope. The area is underlain by relatively thick weathered material.
Plateau: Flat topped and arcuate arc showing definite trends. Comprises thin veneer of soil, which varies from place
to place, scanty vegetation Weathering is also found at some places.
Structural Hills (SH): Structural hills are linear or arcuate hills exhibiting definite trend. These hills are structurally
controlled with complex folding, faulting and crisscrossed by numerous joints/fractures, which facilitate some
infiltration and mostly act as run off zones. These units are found in the northern and eastern parts of the study area.

826
Figure 1: Location Map of North Bangalore Metropolitan Region Figure 2: Geomorphology Map
In the study area, majorly exists Pediplains (3946 Sq. km), Denudational Hills (124.9 Sq.km), Plateau (12.6 Sq.km)
and Structural Hills (24.9 Sq.km). Eastern part of the study area experiences 700-800mm rainfall and Western part
of the study experiences 800-891 mm of rainfall. The rainfall is positively correlating to the Pediment (Inselberg
Complex) as well as Residual hills of the study area. The study reveals more scope for infiltration and leads to take
up ground water recharge very effectively.
b. Trend Analysis
The annual rainfall series in respect of seven and adjacent taluk rain gauge stations data were verified for the
presence of trends by applying the one or more of the below maintained methods.
i. Graphical method of Trend analysis
ii. Auto Correlation
iii. Power Spectrum
iv. Smoothed Power Spectrum
v. Moving Averages (5, 7, 9 and 11 years)
i. Graphical Method of Trend analysis:
In Graphical method by (Fig:3.1 to 3.4), seasonal precipitation for all season‟s viz., Pre-Monsoon, South-
West Monsoon, North-East Monsoon and Annual are presenting increasing in trend of 70 to 80 mm for pre-
monsoon, 350 to 450 mm for south-west monsoon, 200 to 225 mm for north-east monsoon and 700 to
850mm annually.
Figure 3.1: Pre Monsoon Season Figure 3.2: South-West Monsoon Season

827
Figure 3.3: North-East Monsoon Season Figure 3.4: Annual Trend Graph
ii. Auto Correlation:
Autocorrelation is the cross-correlation of a signal with itself. Informally, it is the similarity between
observations as a function of the time lag between them. It is often used in signal processing for analyzing
functions or series of values, such as time domain signals. Autocorrelation coefficients ranges lag 6 to
37years have been determined for seasonal as well as annual Fig: 4.1 to 4.4. Periodicity observed as 37,16
& 6 years (PM), 12, 37 & 16 years (SWM), 8, 18 & 6 years (NEM) and 16, 22 & 8 years (Annual)
respectively.
Figure 4.1: Auto Correlation of Pre-Monsoon Figure 4.2: Auto Correlation of SW
Monsoon
Figure 4.3: Auto Correlation of NE Monsoon Figure 4.4: Annual Auto Correlation
iii. Power Spectrum
The Periodicities have been calculated using Fourier series method and power spectrum plotted in Fig:5.1
to 5.4 indicates that their exists periodicity in the PM as 37, 4 &3 years, SWM as 2, 4 &2 years, NEM as 3,
7 & 2years and Annual as2, 4&2years. The periodicity ranges from 2 to 37 years in this method.

828
Figure 5.1: Power Spectrum of Pre-Monsoon
Figure 5.2: Power Spectrum of SW Monsoon
Figure 5.3: Power Spectrum of NE Monsoon Figure 5.4: Annual Power Spectrum
iv. Smoothed Power Spectrum:
The Periodicities have been calculated using smoothed power spectrum plotted in Fig:6.1 to 6.4 indicate
that exists periodicity in the PM, SWM,NEM and Annual are(3, 4 &3)years, (2, 4&3) years, (3, 3 &7)years
and (4, 2 & 55)years respectively. The periodicity ranges from 2 to 55 years in this method.
Figure 6.1: Smoothed Spectrum Values of Pre-
Monsoon
Figure 6.2: Smoothed Spectrum Values of SW
Monsoon
Figure 6.3: Smoothed Power Spectrum Values of
NE Monsoon
Figure 6.4: Annual Smooth Spectrum Values
Similarity of frequencies is existing for Power Spectrum and Smoothed Power Spectrum Values of Pre
Monsoon, South-West Monsoon, North-East Monsoon and Annually.

829
Table 1: Frequencies obtained by spectral analysis.
Harmonic Value Year Power
AC PS SPS AC PS SPS AC PS SPS
Pre
Monsoon
3 3 31 37 37 3 0.2 114 84.2
7 30 30 16 4 4 0.2 100 80
18 32 32 6 3 3 0.2 90.1 78.5
South-west
Monsoon
9 45 45 12 2 2 0.2 2764 1311
3 26 26 37 4 4 0.2 1715 1237
7 51 37 16 2 3 0.2 1633 1133
North-East
Monsoon
14 32 32 8 3 3 0.1 1159 756
6 15 33 18 7 3 0.1 863 648
18 54 16 6 2 7 0.1 761 628
Annual
7 49 26 16 2 4 0.2 5727 3126
5 26 49 22 4 2 0.2 4419 3001
13 45 2 8 2 55 0.2 3387 2754
AC- Autocorrelation, PS- Power Spectrum, SPS- Smoothed Power Spectrum
v. Moving Average (5, 7, 9 & 11 Years):
Moving average is the important method of understand the periodicity of rainfall. One of the simplest, and
perhaps most common, smoothing technique employed is that of fitting a moving average (e.g., Thompson
and Ibbitt, 1978, Tomlison, 1980b), In the case of simple moving average, all values are weighted equally.

830
Figure 7: Moving Average of 5, 7, 9 and 11yearsfor (a) of Pre-Monsoon, (b) SW-Monsoon,
(c)NE-Monsoon and (d) Annual.
In Pre-Monsoon season the Moving Average of 5, 7, 9 and 11years are 7 to 47 years and followed by South-
West Monsoon season having 13 to 60 years, North-East Monsoon season is 17 to 41 years and annually as
11 to 46 years. Normal moving average for 5, 7, 9 & 11 years of all seasons represents7to 57 years.
Table 2: Season wise Moving average analysis of Rainfall
Years
Pre
Monsoon
South-west
Monsoon
North-East
Monsoon
Annual Average
5 29,7 20,36,25 23,21,18 13,45,25
7 to 57
7 33,12,30 43,15,22 41,23 57,11,26
9 21,47,30 49,13,21 21,20,22 46,13,22
11 33,29 60,22 17,22 44,22
vi. Seasonal and Annual Rainfall
Figure 8.1: Pre-Monsoon Rainfall Map Figure 8.2: South-WestMonsoon Rainfall Map

831
Figure 8.3: North-EastMonsoon Rainfall Map Figure 8.4: Annual Rainfall Map
Table 3: Seasonal and Annual average Rainfall
Sl No Taluk PM SWM NEM ANNUAL
1 Bangalore North 176 490 225 891
2 Devanahalli 147 406 205 759
3 Doddaballapura 134 410 196 740
4 Hosakote 148 385 224 757
5 Nelamangala 166 447 218 831
6 Magadi 173 454 225 852
7 Malur 151 358 214 722
The rainfall maps of all seasons and annual proves the Köppen Classification system of Climatic zones for the study
area as follows:
1. Semi Arid Zone (Western part)
2. Tropical Wet and Dry (Eastern part)
The rainfall of study area experiences 109.7 to 193.1 mm in Pre-Monsoon, 353.7 to 489.9 mm in South-West
Monsoon, 184 to 321.4 mm in North-East Monsoon and 684 to 891 mm annually is presented in Table 3.South-
Western part of the study area is experiences heavy rainfall and North-Eastern part is experiencing least rainfall.
Conclusion
The rainfall maps of all seasons and annual proves the Köppen Classification system of Climatic zones for the study
area. Rainfall being the only source of water in the region, sustenance of the forest depends on amount of rainfall,
water availability in the year. The knowledge of excess or drought in the study area is important from the water
management point of view. Artificial recharge and creation of water bodies in this region may be considered
effectively. From the graphical method indicates the increasing trend of rainfall for Pre-monsoon, South-West
Monsoon, North-East monsoon and annually. The study reveals that the periodicity varies from 5to 57years by
moving average study; 6 to 37 years by autocorrelation study, 2 to 37 years by Fourier spectral analysis and2 to 55
years by smoothing Fourier analysis. Rainfall in the study area is influencing in South-Western part for Pre-
monsoon, South-West Monsoon, North-East monsoon and annually. The southwest and southeast parts of the study
area experience the heavy rainfall whereas the least rainfall areas are the northern parts of the study area.The short
term and long term cyclicity observed in Autocorrelation, power spectrum and Moving Average.
References
Burn, D.H. and M.A. Hag Elnur (2002), Detection of Hydrologic Trends and Variability. Journal of Hydrology
255, 107-122
Goswami, B. N., V. Venugopal, D. Sengupta, M. S. Madhusoodanan, and P.K. Xavier (2006) Increasing trend
of extreme rain events over India in a warming environment Research, 20: 127-136, Science, 314, 1442 – 1444.

832
IshappaMuniyappaRathod ,Aruchamy.S (2010), Spatial Analysis of Rainfall Variation in Coimbatore District
Tamilnadu using GIS. International Journal of Geomatics and Geosciences Volume 1, No 2.
Janki.H.Vyas, Dr.N.J.Shrimali and Mukesh.A.Modi (2012).Trend Analysis of Rainfall of Junagadh District.
Indian Journal of Research Vol.1 (10): 74-75.
R. Stella Maragatham, Trend Analysis of Rainfall data - A Comparative study of Existing methods. 2012 Vol 2 (1)
January- March, pp. 13-18, ISSN 2277-2111(Online).
Robert L Beschta (1982), Moving Averages and Cyclic Patterns. Journal of Hydrology (N.Z) Vol.21, No.2, 148-
151
Sahai AK, Grimn AM, Satyan V, Pant GB (2003) Long lead prediction of Indian summer monsoon rainfall from
global SST Evolution.- Climate Dynamics 20 855-863
Thomson DJ (1982). Spectrum estimation and harmoni
View publication statsView publication stats

Time Series Analysis of Rainfall in North Bangalore Metropolitan Region using Remote Sensing & Geographic Information System Techniques

More Related Content

What's hot (20)

Similar to Time Series Analysis of Rainfall in North Bangalore Metropolitan Region using Remote Sensing & Geographic Information System Techniques (20)

Recently uploaded (20)

Time Series Analysis of Rainfall in North Bangalore Metropolitan Region using Remote Sensing & Geographic Information System Techniques