INTERNATIONAL JOURNAL OF GEOMATICS AND GEOSCIENCES Volume 6, No 3, 2016 © Copyright by the authors - Licensee IPA- Under Creative Commons license 3.0 Research article ISSN 0976 – 4380
Spatio-temporal landuse/landcover (LULC) change analysis of Kolong River basin, Assam, India using Geospatial technologies Minakshi Bora, Dulal C Goswami Department of Environmental Science, Gauhati University, Guwahati-781014, Assam, India
[email protected] ABSTRACT The Kolong River of Nagaon district in Assam, which once used to be a prize possession for the people of the state in general and for the people of Nagaon in particular, is presently gasping on its death-bed because of a ruthless and untenable act perpetrated on it in the name of engineering solution to the increasing flood hazard attributed to it in the aftermath of the great Assam earthquake of 1950. In the backdrop of the above scenario, it is high time we undertake a holistic river restoration programme for the Kolong River based on state-of-the art knowledge and scientific know-how currently available on the subject. Application of remote sensing and GIS, techniques will serve as a basic set of mapping tools for creation of a baseline overview of the river basin. Attempt has been made in this study to map out the status of land use/land cover of Kolong basin with a view to detect changes that had occurred in their status particularly in the built-up area and agricultural land. The landuse/ landcover of the study area have been categorized into six broad classes based on the variation observed in the area. The results showed a significant negative change in total agricultural land with a total decrease from 2216.96 Km2 to 1449.39 Km2 during the year 1967-68 to 2014; while an overall positive change was observed in case of built-up area which has increased from 1069.05 Km2 to 1838.84 Km2 during the study period. Single layer overlay analysis was also performed in order to verify the overall change under each landuse/landcover class. Significant reduction (704.97Km2) in agricultural land to built-up area was verified by the conversion matrix. Keywords: Kolong River, restoration, landuse/landcover, GIS, conversion matrix 1. Introduction The Kolong River of Nagaon district in Assam, which once used to be a prize possession for the people of the state in general and for the people of Nagaon in particular, is presently gasping on its death-bed because of the ruthless and untenable act perpetrated on it in the name of engineering solution to the increasing flood hazard attributed to it in the aftermath of the great Assam earthquake of 1950. During the years preceding 1964, primarily as a consequence of the great Assam earthquake of 1950, this region experienced recurrence of large floods due mainly to increased bed level of the Brahmaputra through massive sedimentation vis-à-vis the level of Kolong, leading thereby to its higher flood levels inundating the low lying adjoining areas like Nagaon town. Mainly as a response to the increasing food hazard faced by the district administrative headquarter i.e. the Nagaon town, an ad hoc flood control measure was undertaken by constructing an earthen embankment, known as Hatimura dyke, across the river’s take-off point near Hatimura in the year 1964. This drastic human intervention had resulted in converting the once free flowing river into a string of alternating dry stretches and stagnant pools during the decades that followed (Bora and Goswami, 2014). Submitted on November 2015 published on February 2016
1676
Spatio-temporal landuse/landcover (LULC) change analysis of Kolong River basin, Assam, India using Geospatial technologies Minakshi Bora, Dulal C Goswami
Although the flood problem has been tackled to some extent by the dyke, but the adverse impacts it had on the entire length of the river from environmental, social and cultural points of view were totally demeaning and destructive. In the subsequent years, the bed of the river started getting filled in with sediments, waste materials and encroached by water hyacinth and other weeds. Earlier this very region was a paradise of nature’s bounty, but due to lack of enough river flow, especially during the dry months of winter, the river dries up almost completely leaving only putrid pools of water here and there (Bora and Goswami, 2015). The river bed has risen and its banks got lowered as a result of deposition of wash-off materials from surrounding areas and extensive growth of weeds like Eichhornia crassipes, Ipomoea aquatic, Ipomoea carnea etc. There has been a massive decline in agricultural productivity in the upstream regions because of lack of sufficient water. The unique aquatic and riparian biodiversity which used to flourish earlier is now almost non-existent or in highly degraded state. In the backdrop of the above scenario, it is high time we undertake a holistic river restoration programme for the Kolong River based on state-of-the art knowledge and scientific knowhow currently available on the subject. Application of remote sensing and GIS, techniques will serve as a basic set of mapping tools for creation of a baseline overview of the river basin. Knowledge of landuse/landcover is important for many planning and management activities and is considered an essential tool for modeling and understanding the earth as a system (Barman and Goswami, 2015). Remotely sensed data especially satellite data can be effectively used in mapping as well as monitoring of temporal changes in land use/land cover of an area (Bansal et al, 2012). Currently the act of LULC mapping is being practiced worldwide. Any observed natural physical features on the earth’s surface are termed as the landcover of that particular land parcel, whereas landuse refers to the addition of any socioeconomic function or anthropogenic activities like residential, agricultural, industrial etc. to a given land parcel ( Longley, 2001). The role of a river system on the overall socio-economic setup of a region is sponsored by the LULC pattern of the area. LULC study is important for hydrologic and watershed modeling (Madhurapperuma et al, 2015). Various techniques of LULC change detection analysis were discussed by Lu et al, (2003). The present study was conducted with a view to analyze and monitor the spatio-temporal land use/land cover (LULC) change patterns of Kolong river basin using multi-temporal spatial datasets for the past 46 years (1967 to 2014). Besides, the study is focused on estimating the conversion rates of each LULC class in both spatial and temporal scale. Spatio-temporal LULC change imparts implications over the socio-economic setup of a given region (Singh and Khanduri, 2011). 2. Study area The Kolong River with a total length of about 230 km is a distributary (Suti in local language) of the Brahmaputra which branches out from it near Jakhalabandha, about 77 Km upstream of Nagaon, and meets it again at Kajalimukh near Guwahati in a joint channel with the Kopili river- a major south bank tributary of Brahmaputra that flows into Kolong near Jagibhakatgaon of Morigaon district. On the way, several smaller streams viz. Diju, Misa, Haria and Digaru meet it. The total basin area of the river is about 5300 Sq. Km and the entire area lies in the northern plain region of Nagaon and Morigaon districts of Assam. The river and its catchment area starting from Hatimura to Kajalimukh are shown in the map given below (Figure 1). The mighty Brahmaputra follows almost west-southwesterly course along the northern part of the basin. The geographical location of the origin point is confined International Journal of Geomatics and Geosciences Volume 6 Issue 3, 2015
1677
Spatio-temporal landuse/landcover (LULC) change analysis of Kolong River basin, Assam, India using Geospatial technologies Minakshi Bora, Dulal C Goswami
to 92º 59’ 38” East and 26º35’21” North, whereas the location of the point of confluence with Brahmaputra is confined to 91º57’51” East and 26º14’41” North.
Figure 1: Study area Major part of the Kolong basin is under alluvial plains interrupted only occasionally by scattered inselbergs of gneiss and granite. Hills of Karbi Anglong and Meghalaya finger into the alluvial plains in the southern bank. The altitude of the hillocks ranges from 60 to 110 meters above MSL. However, the northern bank of Kolong catchment is having more or less a gentle slope with only one hillock viz. Bura Pahar in the extreme east near the river’s mouth. 3. Methodology For the present study, two sets of spatial data viz. SOI toposheets for the year 1967-68 and LANDSAT (ETM+) [path-136; row-42] for the year 2014 were used for analysis based on GIS techniques. A total of nineteen SOI toposheets encompass the entire study area whereas a single satellite image covers the same. Various toposheets were first georeferenced using WGS84 datum and then projected to UTM (Zone46) projection. Likewise, the orthorectified satellite image was also projected to UTM projection after proper radiometric corrections. Six broad LULC classes viz. agricultural land, built-up area, wetland, forest cover, shrub land and open space were categorized and area under each class was digitized separately using ArcGIS software. Area under individual class (in square kilometer) alongwith their percentage coverage out of total geographical was calculated and the statistical results were placed in Table1. For assessing the change detection within each land use/land cover class a single layer overlay analysis was performed using the union tool of ArcGIS software. Selected LULC class of the preceding year (1967) was overlaid upon the same class of the succeeding year (2014) to obtain three separate classes i.e. class under no-change, extended (from-class) and diminished (to-class). Thus, LULC net-change under each category has been obtained and the results are summarized in Table 2. For establishing the LULC category into which or from which a given category had been changed the clip function under overlay analysis was used. Finally, a change detection matrix had been developed (Table 3) based on the results of International Journal of Geomatics and Geosciences Volume 6 Issue 3, 2015
1678
Spatio-temporal landuse/landcover (LULC) change analysis of Kolong River basin, Assam, India using Geospatial technologies Minakshi Bora, Dulal C Goswami
overlay analysis. Finally, in order to determine the extent and rate of change in the land cover dynamics in the region, following variables were computed. 1. Total area (TA) 2. Changed area (CA) 3. Change extent (CE) 4. Annual rate of change (CR) These variables can be described by the following formula: CA= TA (t2) – TA (t1) CE=CA /TA (t1) CR=CE / (t2 – t1); Where t1 and t2 are the beginning and ending time of the land cover studies conducted i.e. 1967 and 2014 respectively. 4. Results and discussion Using the approaches adopted in the methodologies, LULC maps were generated for the two studied years (Figure 2 and Figure 3) and area statistics and change statistics were computed. Area under individual class and change analysis for the two years were summarized in Table 1.
Figure 2: LULC Map of Kolong Basin (1967-68)
Figure 3: LULC Map of Kolong Basin (2014) International Journal of Geomatics and Geosciences Volume 6 Issue 3, 2015
1679
Spatio-temporal landuse/landcover (LULC) change analysis of Kolong River basin, Assam, India using Geospatial technologies Minakshi Bora, Dulal C Goswami
Table 1: Areas and percentages of LULC classes for the period 1967–2014 Sl. No
LULC category
1967-68
2014
Area(Km2)
Area (%)
Area(Km2)
Area (%)
1
Wetland
284.16
5.345
254.076
4.869
2
Agricultural land
2216.96
41.703
1449.39
27.773
3
Built-up area
1069.05
20.11
1838.846
35.235
4
Forest cover
1407.58
26.478
1314.366
25.185
5
Shrub/grassland
283.135
5.326
347.833
6.665
6
Open space
55.23
1.039
14.24
0.273
The analysis of Land use/land cover, indicated substantial changes. Out of the total geographic area, larger proportion was under agricultural land (41.7%) during the study period 1967-68, whereas during 2014 larger portion of the total geographic area accounts for built-up area (35.24%) (Figure 4).
International Journal of Geomatics and Geosciences Volume 6 Issue 3, 2015
1680
Spatio-temporal landuse/landcover (LULC) change analysis of Kolong River basin, Assam, India using Geospatial technologies Minakshi Bora, Dulal C Goswami
which or from which a given category had been changed. Finally, a change detection matrix had been developed (Table 3) based on the results of overlay analysis. Table 2: Net-change under various LULC classes (1967 to 2014) Landuse category
No change (Km2)
Extended (Km2)
Diminished (Km2)
Net-change (gain – loss)
Wetland Agricultural land Built-up area
122.49
131.33
161.67
–30.34
1172.04
278.06
1060.29
–767.57
856.341
982.52
212.71
769.81
Forest cover
1271.78
42.58
135.80
–93.22
Shrub/grassland
100.53
246.70
183.05
63.65
Open space
2.19
11.74
53.04
–41.3
a
b
c
d
International Journal of Geomatics and Geosciences Volume 6 Issue 3, 2015
1681
Spatio-temporal landuse/landcover (LULC) change analysis of Kolong River basin, Assam, India using Geospatial technologies Minakshi Bora, Dulal C Goswami
e
f
Figure 5 (a, b, c, d, e and f): LULC change detection map of Kolong basin(1967 to 2014) The study area exhibits a significant LULC change during the study period. To further evaluate the results of land cover conversion, matrices of land cover changes from 1967 to 2014 were created (Table 3). Table 3: LULC Change detection matrix for 1967-2014 (Area in Km2) Agricultural Built-up Forest Shrub/ Open Change from Wetland Land area Cover grassland Space Wetland 122.49 72.77 52.13 2.05 29.22 0.25 Agricultural 105.11 1171.46 704.97 16.15 154.65 5.9 land Built-up area 11.5 111.7 856.34 9.57 18.87 2.58 Forest cover 1.23 30.1 72.85 1241.78 31.2 0.256 Shrub/grassland 12.95 42.06 95.12 10.6 100.53 1.8 Open space 2.53 18.029 19.3 2.19 3.66 2.19 N.B: Data in bold represent the unchanged fractions within each class The LULC change detection matrix revealed that majority of the agricultural land which experienced an overall negative change has been converted into built-up area (i.e. 704.97 Km2). Wetlands which are not only important ecological features but which also reflects the hydrological health of a given area also showed a negative trend (Figure 4) and majority of changed area under wetland (72.77Km2) had been converted into agricultural land. Forest cover representing the biodiversity of the study area also represents a net-negative change and the LULC change detection matrix revealed the fact that built-up area had expanded at the expense of forest cover and a total of 72.85 Km2 land had been added under built-up area from forest cover. Similarly, maximum area under shrub land and open space were also found to be converted into built-up area (Table 3). However, forest cover represented the LULC class having maximum unchanged fraction followed by agricultural land, built-up area, wetland, shrub land and open space respectively (Table 3). In the present study an attempt had also been made to determine the extent and rate of LULC conversion. The rate of change was as high as 1.53% for built-up area. On the contrary, wetland, agricultural land, forest cover and open space showed an average annual reduction International Journal of Geomatics and Geosciences Volume 6 Issue 3, 2015
1682
Spatio-temporal landuse/landcover (LULC) change analysis of Kolong River basin, Assam, India using Geospatial technologies Minakshi Bora, Dulal C Goswami
of change approximately 0.23%, 0.74%, 0.14% and 1.58% respectively. The rate of change in wetland and forest cover was insignificant during the study period. The highest extent of change was observed for open space representing 74.21% of change within 47 years difference (Table 4). Table 4: overall amount, extent and rate of LULC change (1967-2014) 1967 to 2014 LULC category 2 Change (∆/Km ) Extent (%) Rate of ∆(%/year) Wetland -30.084 -10.58 -0.23 Agricultural land -767.57 -34.62 -0.74 Built-up area +769.8 +72 +1.53 Forest cover -93.214 -6.6 -0.14 Shrub/grassland +64.698 +22.85 +0.5 Open space -40.99 -74.21 -1.58 5. Conclusion It is believed that the present study will help to contribute towards sustainable landuse planning and management towards protection of extremely rich biodiversity of the Kolong basin. On account of increasing population and eco-hydrological changes the area is facing land use changes. The alarming fact that came into focus out of the present study is the high annual rate of change observed under built-up area and that maximum conversion into builtup area is contributed by agricultural land. The demeaning condition of the main river system i.e. the Kolong River along with increased demographic pressure is the possible factor behind the present LULC conversion pattern. Acknowledgement The authors’ thanks are due to the INSPIRE program sponsored by the Department of Science and technology (DST), Govt. of India, for extending financial assistance for the present research. 6. References 1. Bansal A., Karwariya S., Goyal S., (2012), Change Detection in L use / L cover in Sewan Watershed Using Remote Sensing GIS Technique, International Journal of Advances in Remote Sensing GIS, 1(2), pp 208-217. 2. Barman P., Goswami D.C., 2015, Luse/Lcover mapping of Dhansiri (South) River Basin, Assam using remote sensing GIS techniques, International Journal of Geomatics Geosciences, 5(3), pp 474-481. 3. Bora M., Goswami D.C., (2014), Study for Restoration using field survey Geoinformatics of the Kolong River, Assam, India, Journal of Environmental Research Development, 8(4), pp 997-1004. 4. Bora M., Goswami D.C., (2015), A study on seasonal temporal variation in physicochemical hydrological characteristics of River Kolong at Nagaon Town, Assam, India, Archives of applied science research, 7(5), pp 110-117. International Journal of Geomatics and Geosciences Volume 6 Issue 3, 2015
1683
Spatio-temporal landuse/landcover (LULC) change analysis of Kolong River basin, Assam, India using Geospatial technologies Minakshi Bora, Dulal C Goswami
5. Longley P., Donnay J., Bransley M., (2001), Remote sensing urban analysis, London, Taylor Francis. 6. Lu, D., Mausel, P., Brondizio, E., Moran, E., (2003), Change detection techniques, International Journal of Remote sensing, 25, pp 2365–2407. 7. Madhurapperuma B., Rozario P., Oduor P. Kotchman L., (2015), L-use l-cover change detection in Pipestem Creek watershed, North Dakota, International Journal of Geomatics and Geosciences, 5(3), pp 416-426. 8. Singh, A., (1989), Review Article: Digital Change Detection Techniques using Remotely-Sensed Data, International Journal of Remote Sensing, 10(6), pp 989- 1003. 9. Singh P., Khuri K., (2011), L use L cover change detection through Remote Sensing and GIS Technology: Case study of Pathankot Dhar Kalan Tehsils, Punjab, International Journal of Geomatics Geosciences, 1(4), pp 839-846.
International Journal of Geomatics and Geosciences Volume 6 Issue 3, 2015
1684