Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 1

ISS N 2229-5518

Land Use and Land cover for one Decade in Coimbatore Dist Using Historical and Recent High Resolution Satellite Data

M.Renuka Devi, Lt.Dr.S.Santhosh Baboo

Abs tract--The aw areness of landuse / land cover assessment is very important to understanding natura l resources, their utilization, conservation and manage ment. In recent years remote sensing and Geographical Inf ormation System have gained importance as vit al tools in the analysis of temporal data at the district and citylevel. The present study evaluates the eff ectiveness of high-resolution satellite data and computer aided GIS techniques in assessing landuse / land cover change detection f or the period 1990 to 2000 w ithin the study area, Coimbatore District. This paper describes assessment of the land use and land cover changes in the Coimbatore District f or one decay. IRS IC images of 1990 and 2000 w ere analyzed using Erdas Imagine sof tw are and ArcGIS. A total of f ive broad landuse and land cover classes w ere identif ied. These w ere crop land ,Barren land,f orest , w ater bodies and built up land. This study identif ied population grow th, built up land and lack of proper education as causes of the changes in land use and land cover in the Coimbatore area.

Ke ywords : land cover change, land use change, IRS IC, re mote sensing and Geographical Inf ormation System

INTRODUCTION

—————————— ——————————
The r ecent availability of high-r esolution satellite imagery has led to incr eased inter est in the use of satellite data for lar ge scale mapping applications and detailed land use assessments. This gr owing inter est not only emanates fr om the fact that satellites pr ovide a synoptic coverage, have a high r epetitive cycle, and carry multispectral band sensors that pr ovide infor mation beyond the or dinary ability of the human eye, but also because they offer a costeffective source of data that enables timely detection of changes to the landuse and landcover , the monitoring and mapping of urban development, assessment of defor estation extents, evaluation of post fir e vegetation r ecovery, the r evision of topogr aphic maps among numer ous other envir onmental assessments.
Land use and land cover information ar e important for
several planning and management activities concer ned with the sur face of the earth because it constitutes key envir onmental information for many scientific, r esour ce management and policy pur poses, as w ell as for a range of human activities.

1 M.Renuka Devi,Ph.D Pa rt Time Resea rch Scholor

Bha ra thia r University,Coimba tore.

Assista nt Professor in MCA Depa rtment

Sree Sa ra swa thi Thayaga raja College,Pollachi

Email-re nug a.srk@g mail.co m, 94875 90190

1 Lt.Dr.S.Sa nthosh Ba boo, Reader, Postgradua te a nd

Resea rch depa rtment of Computer Science,

Dwa raka Doss Goverdhan Doss Va ishnav College, Chenna i.

Email- santho sh2001@sify.co m

An accurate knowledge of land use and land cover featur es r epr esents the foundation for land classification and management. Ther efor e a wide r an ge of scientists and practitioners, including earth systems scientists, land and water manager s as w ell as urban planners seek information on the location, distr ibution, type and magnitude of land use and land cover change.
Vegetation changes ar e often the r esult of anthr opogenic
pr essur e (e.g. population gr owth) and natur al factors such as variability in climate. Due to incr easing population growth rates, ther e have been incr easing rates of conver sion of for est and built up land in developing economies all over the w or ld. The degr adation of for est
have impact on catchment pr ocesses and biochemical cycles and leads to soil er osion and water shortage not only in the r egions immediately affected by defor estation, but also in r easonably distant ar eas .This means that pr oblems posed by land use and land cover change are numer ous and have serious consequences.
Ther efor e, the spatial dimensions of land use and land cover needs to be known at all times so that policy - makers and scientists will be amply equipped to take decisions. The most impor tant thing is the changing patter n of land use and land cover r eflect changing economic and social conditions.

2. STUDY AR EA

IJSER © 2012 http :// www.ijser.org

Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 2

ISS N 2229-5518

Coimbator e, Distr ict is situated on the banks of river Noyyal betw een 11° 00' of north latitude and 77° 00' of East longitude. The total ar ea of Coimbator e district is 254 squar e km. Coimbator e is located at an elevation of about
398 meter s. The mean maximum and minimum
temperatur es dur ing summer and w inter var ies between
35°C to 18°C. Highest temperatur e ever r ecor ded is 41 °C and low est is 12 °C. Coimbator e is situated in the extr eme w est of Tamil Nadu, near the state of Kerala. It is
surr ounded by mountains on the west, with r eser ve for ests and the (Nilgir i Biospher e Reserve) on the northern side. The easter n side of the distr ict, including the city is pr edominantly dry. The entir e w ester n and northern part of the distr ict bor ders the W estern Ghats with the Nilgiri biospher e as w ell as the Anaimalai and Munnar ranges. It is the third lar gest distr ict of Tamil Nadu. This distr ict is known as the Manchester of South India and is known for its textile factor ies, engineering firms, automobile parts manufactur er s, health care facilities, educational institutions, and hospitality industr ies. The hill stat ions of Ooty, Coonnor and Valparai ar e close to the city making it a good tour ist attraction thr oughout the year . The district is situated on the banks of the Noyyal River and is close to the Sir uvani Water falls. This distr ict contains four main r eservoir that is Aliar ,Thirumurthi ,Amar avathi and Solaiar .

3. D AT A AND IM AG ERY U SED

IRS IC and IRS IA imager ies w er e collected fr om Water
Resear ch Institution,Tharamani, Chennai.Topo Sheet 58
A,58E,58B and 58 F wer e collected fr om Survey of
India,Map Sales office, Chennai.

4. IM AGE PROCESSING

IRS IC images of scene of year 1990 and 2000 showing the r oads, towns, and drainage systems w ere used for the study. Remote sensing softwar es: Er das Imagine version 9.2 and Ar cGIS ver sion 9.2 w er e used for the pr ocessing of the images. The raw satellite image was converted fr om Tag Image file format (Tiff) to img format using Er das in order to be compatible with other Er das Imagine files. The layer s wer e stacked and sub-set to delineate the catchment ar ea for classification. The UTM Zone 30N Coordinate on the WGS 84 w as used to geocode the imported image. This was followed by geor efer encing using the Traver se Mer cator pr oj ection with r efer ence units in meters to allow compatible positioning of other themes such as r oads, towns and
drainage which w er e alr eady digitized in that format. Then the digitized map show ing the r oads, towns, drainage and the outline of the r eservoir was over laid on the geor efer enced image. Band combination of r ed, blue and gr een was used to display the r aw images in standar d colour composites. The spectr al band combination for displaying images often var ies w ith differ ent applications (Tr otter , 1998). This was necessary for the visual inter pr etation of the images. A band combination of r ed, blue and gr een (RGB) is often used to display images in standard colour composites for land use and vegetation mapping (Tr otter , 1998). In this study, the IRS IA 2000 images w er e displayed in a band combination of 3,2 and 1 (r ed, blue and gr een), which is standar d for visual inter pr etation of land use and land cover mapping in the tr opics (Prakash and Gupta 1998; Tr otter , 1998).The 1990 imagery is visually interperted by using Ar cGIS (Version 9.1)

4.1 Land Cover Cla ssification

The super vised classification method was used to classify the images into the var ious land cover categor ies. The maximumlikelyhood supervised classification method is applied for gr ouping the pixel in IRS IC 2000 imagery. . The selection of appr opr iate training ar eas is based on the analyst's familiarity with the geographical ar ea and their know ledge of the actual sur face cover types pr esent in the image. Thus, the analyst "supervises" the categor ization of a set of specific classes. The numer ical information in all spectr al bands for the pixels compr ising these ar eas ar e used to "train" the computer to r ecognize spectrally similar ar eas for each class.
Training Pixels
Training fields ar e ar eas of known identity
delineated on the digital image, usually by specifying the corner points of a r ectangular or polygonal ar ea using line and column numbers within the coordinate system of the digital image. The analyst must, of course, know the corr ect class for each ar ea. Usually the analyst begins by assembling maps and aerial photographs of the ar ea to be classified. Specific training ar eas ar e identified for each
informational categor y follow ing the guidelines outlined below. The obj ective is to identify a set of pixels that accur ately r epr esents spectral variation pr esent within each information r egion (Fig. 1a,1b ).
Training ar ea pr ocess is called signatur e cr eation is shown in the following figur e5a,5b. In the figur e some of the classes like cr opland, water body, barr en land and hills wer e chosen as tr aining ar ea. In this pr ocess the r ed pixels ar e trained as cr opland, block pixels ar e trained as water body or tank, the ash gr een pixels ar e tr ained as

IJSER © 2012 http :// www.ijser.org

Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 3

ISS N 2229-5518


barr en land and dar k gr ay pixels selected as Hills continue this pr ocess accor ding to our classification scheme.

Cropland Waterbody


Figure 1a .Training the red pixels as cropland and block texture as w aterbody

Land Hills

Figure 1b. Training the pixel as Barren Land and Hills

The statistics of the var ious classes w er e gener ated using the Er das Imagine pr ogramme.Finally maps w ere composed, using pr ogr amme and the maps w ere validated in the field to assess its accur acy. This was conducted thr ough field visit to define how closely the classification agr ees with the actual field situation. It involved the selection of samples of identified locations on the map, w hich w er e then checked in the field.

4.2 Change Detection

Ther e ar e lots methods ar e available to find out the change detection in land. The most fr equently used land change detection methods includes i) image over lay ii) classification compar isons of land cover statist ics iii) change vector analysis iv) pr incipal component analysis and v) image r ationing and vi) the differ encing of normalized differ ence vegetation index (NDVI) (Duadze,
2004). This r esearch used classification comparison of
land cover statistics. This method was adopted because the study to find out the changes in the ar eas of the various land cover categor ies. Using the post -
classification pr ocedur e, the ar ea statistic for each of the land cover classes was derived from the classifications of the images for each date (1990 and 2000) separately, using functions in the Er das Imagine softwar e. The ar eas cover ed by each land cover type for the one decay w ere
compar ed. Then the dir ections of the changes in each land cover type 1990 and 2000 w er e determined.

5.RESULT DISCU SSION

5.1 Re sults o f land cover classifi cation

Ther e ar e totally seven categories wer e identified and classified in this study. They ar e water body, for est, settlement, cr opland, barr en land, fallow land and others (doesn’t include ab ove categories). The classification of these categor ies w er e shown in the Figur e 2

Figure 2.Land Cover Classif ication 1990 and 2000 Imageries

5.2. Land cover changes for one decay

Table 1 shows the changes in the various land use and land cover categor ies (in sq.km and per centages) during the per iods between 1990 to 2000.In the table (+) denotes that the percentage incr eased and (-) denotes that the percentage decr eased.

TABLE 1.

LAND COVER CHANGES IN THE PERIOD OF 1990 TO

2000

IJSER © 2012 http :// www.ijser.org

Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 4

ISS N 2229-5518

Classificatio

n

LANDUSE

1990

Sq.km

LANDU

SE 2000

Sq.km

CHANGE DURING

1990-2000

Classificatio

n

LANDUSE

1990

Sq.km

LANDU

SE 2000

Sq.km

Sq.m

Percenta

ge%

Wate rbo die s

1618.130

1573.68

44.450

0.51(-)

Se ttle me nt

112.161

633.02

520.859

5.94(+)

Fo re st

1370.413

1522.02

151.607

1.73(+)

Cro p Land

912.531

1141.09

228.559

2.61(+)

Barre n Land

4563.534

2346.77

2216.764

25.29(-)

Fallo w Land

163.38

533.75

370.370

4.22(+)

Othe rs

26.334

1016.15

989.816

11.29(+)

Total

8766.483

8766.48

5.3. Cau ses o f the land cover changes

In the period of one decay settlement was incr eased up to

12% this pr oves the population gr owth. Population gr owth is the basic factor for envir onmental change, because this is key factor for all development especially in developing countr ies like India. In the below figur e 3 w e can see the changes in settlement in 2000. Fortunately the per centage of water body was only 1% decr eased. In the total ar ea 8766.483 sq.km 3% of for est w as incr eased. The per centage of fallow land was incr eased and barr en land was decr eased. Mainly the cropland was incr eased upto 5%. This is shown in the figur e 3.

Change De te ction for one De cay

Waterbodies

Settlement

Figure 4. Show ing Changes betw een Settlement betw een 1990 to

2000

In the figur e 4 compar e with left side imager y with r ight imagery settlement(sky blue textur e) ,cr opland( Red color )and waterbodies(Block color ) w er e incr eased in the one decay.

6.CONCLUS IONS

In the analysis of IRS IC Images 1990 and 2000 exposed that land use and land cover of Coimbator e settlement has changed over the one decay. Seven classes of land use

11.29

4.22

0.51 5.94

1.73

2.61

25.29

Forest Crop Land Barren Land Fallow Land Others

Total

and land cover w er e identified and mapped using above
imagery. The used classes w ere waterbodies,for est,settlement,cr op land fallow land,barr en land others.generally if settlement was incr eased then gr owth of populat ion w ill incr ease. This is pr oved by incr ement in the Coimbator e settlement. Ther efor e based on the population gr owth the land use

Figure 3.Change Detection Chart f or one Decay

and land cover tr end is changed.But only few per centage of water was decr eased.

REFERENCES

[1] Be rnard, A. C., Wilkinso n, G. G. and Kane llo po ulos, I., 1997. Training strate g ie s fo r ne ural ne two rk so ft c lassific atio n o f re mo tely se nse d imag e ry.int.j.re mo te se nsing , 1997, Vo l.18, 1851 -1856.

[2] BWEP,2005.Bare kese Wate r Expansio n Projec t. Draft Re po rt

(ve rsio n), 9R3818.A0/R/ACT/ Nijm,14 Se pt.2005.

[3] Che ng , G.W., 1999. Fo rest c hange : hy dro log ic al e ffec ts in the uppe r Yang tze rive r v alley . Ambio 28, 457–459.

[4] Co sta, M. H., Bo tta, A. and Cardille , J. A., 2003. Effec ts o f Large - Sc ale Changes in Land Cove r and Climate V ariability in the Disc harge o f the Toc antins Rive r, Ame ric an Geo phy sic al Unio n, Fall Mee ting 2002, abstrac t #B22C-0765

IJSER © 2012 http :// www.ijser.org

Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 5

ISS N 2229-5518

[5] Co ng alto n, R. G., 1991. A rev ie w o f asse ssing the acc uracy o f c lassific atio ns o f re mo te ly se nse d data. Re mo te Se nsing o f Env iro nme nt, 37, 35 -46

[6] Dai, X. and Kho raran, S ., 1998. A hie rac hic al

me tho do logy frame wo rk fo r multiso urc e data fusio n in vege tatio n

c lassific atio n.int.j.re mo te se nsing ,1998, Vol.19, no .18, 3697 -3701

[7] Damizade h, M., S ag hafian, B. and Gre ske, A., 2000. S tudy ing

vege tatio n re spo nses and rainfal l re latio nship base d o n NOAA -

AV HRR images

[8] De Fries, R. and Be lward, A. S ., 2000. Glo bal and re g io nal land

cove r c harac te rizatio n fro m sate llite data; an intro duc tio n to the S pec ial Issue .int.j.re mo te se nsing, 2000, vo l.21, no . 6&7, 1083-1092. [9] De Mo raes, J. F. L., Seyle r, F., Ce rri, C. C. and Vo lko ff, B., 1998. Land cove r mapping and c arbo n pools estimste s in Ro ndo nia, Brazil.int.j.re mo te se nsing , 1998, Vo l.19, no. 5, 921 -934

[10] ERDAS. 1999. Fie ld g uide. Fifth e ditio n. ERDAS Inc . Bufo rd

Hig hwas, NE, Atlanta, Geo rg ia. USA, 75(8), 2430 -2437.

[11] Duadze S.E.K. 2004. Land-use and land-cove r study o f the

S av annah eco syste m in the Uppe r West reg io n (Ghana) using re mo te se nsing . ZEF Bo nn, Unive rsity o f Bo nn, Ge rmany FAO.

1999. Ag rostat.http/www.FAO.o rg

[14] Gue rra, F., Puig , H. and Chaune , R., 1998. The fo rest-sav annah

dy namic s fro m multi-data LANDSAT-TM data in S ie rra Parima,

Ve nezue la.int.j.re mo te se nsing , 1998, Vo l.19, no.11, 2061 -2075

[15] He lme r, E. H., Bro wn, S. and Co he n, W. B. 2000. Mapping

Mo ntane tro pic al fo rest successio n stage and land use with mul ti -

dae LANDSAT image ry.int.j.re mo te se nsing , 2000, Vo l. 21, no . 11,

2163-2183

[16] Hill, R. A. 1999. Imag e seg me ntatio n fo r humid tro pic al fo rest c lassific atio n in LANDSAT TM data.int.j.re mo te se nsing , 1999, Vo l.

20, no . 5, 1039-1044

[17] Ho ug hto n, R. A., 1995. Land-use c hange s and the c arbo n cyc le. Glo bal Change Bio logy. 1, 275 -287

[18] IGBP-IHDP, 1999. Land use and land cove r c hange imple me ntatio n strate gy. IGBP Re po rt 48 and IHS P Re po rt 10. IGBP Sec re tariat, S toc kho lm, Swe de n. Pp287

[19] Jane tos, A. C. and Justic e , C. O., 2000. Land cove r g lo bal pro duc tiv ity : a me asure me nt strategy fo r the NASA prog ramme .int.j.re mo te se nsing, 2000, Vol.21, no .6&7, 1491 -1512

[20] Lile sand, T. M. and Ke ife r, R. W., 1994. Re mo te se nsing and

Imag e Inte rpre tatio n. Jo hn Wiley and so ns. Pp750Mas, J. F., 1999.

Mo nito ring land c ove r c hanges; a co mpariso n o f c hange de tec tio n tec hnique s .int. j. re mo te se nsing, 1999, vo l. 20, no . 1, 139 -152

[21] Mille r,A.B., Bry ant,E.S . and Birnie , R. W.,1998. An analy sis o f land cove r c hange s in the No rthe rn Fo rest o f Ne w Eng land using multite mpo ral LANDSAT MSS data. int.j.re mo te se nsing , 1998, Vo l.

19, no . 2, 215-265

[22] NPs/NBS Vege tatio n Mapping , 2002. USGS -Vege tatio n

Mapping Prog ram. http:// bio logy.usgs.gov/nps/aa/sec t1.html

[23] E. Bo akye, S . N. Odai, K. A. Adje i, F. O. Anno r. Landsat Imag es fo r Asse ssme nt o f the Impac t o f Land Use and Land Cove r Changes o n the Bare ke se Catc hme nt in Ghana, Euro pe an Jo urnal o f Sc ie ntific Re se arc h ISS N 1450 -216X Vo l.22 No.2 (2008), pp.269 -278

BIBLIOGRAPHY


Mrs. M.Renuka Devi , has near ly 10 years of post graduate teaching experience in Computer Science. She has indulged in tr aining the post graduate students to complete r eal time pr oj ects and also guides r esearch scholars in Computer Science. Curr ently she is w or king as
Asst.Prof in the Department of MCA at Sr ee Saraswathi
Thyagaraj a College (Autonomous),and An ISO 9001
Certified / NAAC Accr edited Institution, Pollachi, Coimbator e (Dt), Tamil Nadu,India.

Lt.Dr .S.Santhosh Baboo, has ar ound Seventeen year s of postgraduate teaching exper ience in Computer Science, which includes Six years of administrative experience. He is a member , boar d of studies, in several autonomous colleges, and designs the curr iculum of under gr aduate and postgr aduate pr ogrammes. He is a consultant for starting new courses, setting up computer labs, and r ecruiting lectur ers for many colleges. Equipped with a Masters degr ee in Computer Science and a Doctor ate in Computer Science, he is a visiting faculty to IT companies. It is customary to see him at several national/inter national confer ences and training pr ogrammes, both as a participant and as a r esour ce person. He has been keenly involved in or ganizing training pr ogrammes for students and faculty members. His good r apport with the IT companies has been instrumental in on/off campus interviews, and has helped the post graduate students to get r eal time proj ects. He has also guided many such live pr oj ects. Lt.Dr . Santhosh Baboo has author ed a commendable number of r esearch papers in international/national
Confer ence/j our nals and also guides
r esearch scholar s in Computer Science. Curr ently he is Reader in the Postgraduate and Research department of Computer Science at Dwaraka Doss Gover dhan Doss Vaishnav College
(accr edited at ‘A’ gr ade by NAAC),
one of the pr emier institutions in Chennai.

IJSER © 2012 http :// www.ijser.org