Sudhir Kumar Singh
Assistant Professor
Department of Atmospheric & Ocean Science, University of Allahabad, Allahabad, India
[email protected]
Abstract
Multidate remotely sensed data can be used for monitoring land use land cover changes very effectively. This study reveals and quantifies the change over then periods of years by using remotely sensed data and integrating this satellite data with ancillary data. The change analysis was performed by post classification comparison method, comparing the data of two different sensors (Landsat TM and IRS 1C LISS III), at different time periods (years 1989, 2000 and 2005). The results showed that there was rapid change in land cover/land use due to increase in population. It was found that there was a phenomenal change in the built-up area in watersheds, loss of forest cover and change in agriculture land. There is a great need for sustainable management of resources to maximise benefits of societal resources.
Keywords: Remote sensing; satellite data; sensor; soil organic carbon; India; Multi date
Introduction
Land use is the manner in which human beings employ land and its resources including agriculture, urban development, grazing, logging and mining. In contrast, land cover describes the physical state of the land surface, which includes cropland, forests, wetlands, pasture, roads, and urban areas (Jaisawal et al., 1999 and Mukherjee et. al., 2009). The term land cover originally referred to the kind and state of vegetation, such as forest or grass cover, but it has broadened in subsequent usage to include human structures such as buildings, pavements and other aspects of natural environment, including soil type, biodiversity, surface water and groundwater (Chen et al., 2008 and Meyer, 1995). Local land use land cover changes, ranging from losses of wetlands, productive lands and biodiversity to expansion of croplands at the expense of forests across the world are one of the most important human induced disturbance that contribute to global environmental and climate change (Wali et. al., 1999, Evrendilek et. al., 2002 and Kalnay et. al., 2003). Currently, land cover/land use changes due to human disturbance are a net source of carbon emissions to the atmosphere. Change in land use and land cover affects exchanges of energy, water, and momentum between biosphere and atmosphere, as has been demonstrated by experiments which quantify the effects of changes in land use land cover on climate. The worldwide trajectory of land use and land cover changes over the last 300 years has been to decrease the area of forests and increase the area of agricultural lands (Houghton et. al., 2003). Change in land cover/land use also enhances soil erosion, creates strong environmental impacts and high economic costs by its effect on agricultural production, infrastructure and water quality (Lal, 1998 and Pimentel et. al., 1995). Remote sensing has proven to be a cost effective tool for studying changes (Jensen, 1996). Remote sensing techniques can provide the requisite information within very short time at less cost and efforts (Gautam et. al., 1983 and Gautam et.al., 1985). It also provides synoptic coverage of areas of interest and facilitates optimal assessment and monitoring capabilities. These features make remote sensing an optimal tool for this type of study (Alexander et. al., 1973). Remote sensing provides data that are available in time series to study the dynamics of the area over long periods, which bridges mapping and monitoring gaps in understanding the environmental change (Mukherjee, 1998 and Mukherjee 2004). These specifications of remote sensing can therefore greatly contribute to local, regional as well as global mapping and monitoring of change in land cover/land use (LULC) (King et. al., 1993 and Siakeu et. al., 2000)]. The objective of this study was to analyse LULC changes between 1972–1990 and 1990-2005 occurred due to industrialisation, urbanisation and population growth and other developmental activities, for Rupnagar district of Punjab, India, using Landsat Thematic Mapper (TM) and IRS-1C data.
Study area
The study area lies in northeastern part of Punjab, between 30 0 34’ to 31 0 26’N latitude and 76 0 17’ to 76 052’ E longitude. The geographical area of Rupnagar is 2,117 square kilometres. District Rupnagar in Punjab State is a part of kandi belt of Himalayas and alluvial plain of the river Sutlej. Physiographically, there are four main units, namely Shiwalik hills, valleys, piedmont plain and alluvial flood plain. The mean annual rainfall in the district is about 862 mm and the major portion of it is received during the monsoon season, with few showers during winters. The peak temperature of 38-400 C is observed during the month of June. January is the coldest month with mean monthly temperature of 3-40 C. The climate in the Rupnagar study region is semi arid subtropical. Geographically, the study region has been divided into two classes – Precambrian formations and recent formations. Geomorphology is associated with topographic landforms, which in turn are related with runoff and infiltration. The significant geomorphic units identified, based on their image characteristics, include residual hill complex, residual hill, residual mound, pediment, valley fills, valley flat, flood plain and alluvial plain.
Figure1. Showing the Study Area, Rupnagar, Punjab, India
Figure 2.a The classified image of the year 1989
Figure 2.b the classified image of year 2000
Figure 2.c the unsupervised classified IRS IC LISS image of the year 2005 showing various classes mainly the study area during that time having fallow land.
Figure3. Bar diagram showing the percentage of land cover/land use change in the study region during the year from 1989, 2000 and 2005.
Material and Methods
Data Processing. Satellite data can be divided on the basis of spatial resolution in three forms such as coarse (>250 m), medium (>80 m) and fine (<80 m). In this study, we used the medium and fine spatial resolution satellite data. Successful utilisation of remotely sensed data for land cover/land use mapping and monitoring requires careful selection of an appropriate data set and image processing technique(s). Digital change detection is the process of determining and /or describing changes in land cover/land use properties based on co-registered multi temporal remote sensing data. Numerous researchers have addressed the problem of accurately monitoring land cover/land use change in a wide variety of environments with a high degree of success (Muchoney et. al., 1994, Singh, 1989 and Chen et., al., 2001).
The simplest taxonomy separates land cover/land use changes that are categorical versus those that are continuous (Abuelgasim et. al., 1999). Categorical changes in time, also known as post classification comparison, occur between suites of thematic land cover/land use categories. The second taxonomy of change is continuous; it is also known as pre classification enhancement, where changes occur in the amount or concentration of some attribute of the urban/suburban or natural landscape that can be continuously measured (Coppin et.al., 1996).
In our study, we followed the simplest taxonomy method. Once the choice of change detection taxonomy is determined, decisions on data processing requirements can be made. Requirements include geometric/radiometric corrections, data normalisation, change enhancement, image classification and accuracy assessment (Lunetta et. al., 1998).
Geometric correction. Accurate per pixel registration of multi-temporal remote sensing data is essential for change detection since potential exists for registration errors to be interpreted as land cover/land use change leading to an overestimation of actual change (Stow et. al., 1999). Geometric registration is required to remove or reduce the effects of non-systematic or random distortions present in remote sensing data. These distortions can be corrected by developing a model to tie per pixel image features to specific per pixel ground features {i.e. GCP (ground control points)} where geographic coordinates are known from accurate reference maps/image/toposheets or GPS data (Kardoulas et. al., 1996).
Geometric registration error between two images is expressed in terms of an acceptable total root mean square error (RMSE). Several authors recommend a maximum tolerance RMSE value of <0.5 pixels (Gautam et. al., 1983), but others have identified acceptable RMSE values ranging from >0.2 pixels to <0.1 pixels depending upon the type of change being investigated (Townshend et. al., 1992). The images were geometrically corrected and geo-coded to the projection Lambert Conformal Conic, datum Everest India 1830, co-ordinate system, using Survey of India (SOI) toposheets at scale 1:50,000. Ground control points were spatially re-samples the images, using a nearest neighbour algorithm. The transformation had an RMSE of between 0.4 and 0.6 pixel, indicating the accuracy of image rectification within one pixel.
Image classification. Image classification applies to both post-classification and pre-classification change detection approaches and can be performed using either supervised or unsupervised approaches. In supervised classification, calibration data must be sufficiently sampled from appropriate areas and in unsupervised classification, an algorithm is chosen that will take remotely sensed image data set and find a pre-specified number of statistical clusters in measurement space [26]. Although these clusters must then be assigned to classes of land cover/land use, this method can be used without having prior knowledge of ground cover in the study site.
The acquired satellite images of different time periods were classified in software ERDAS 8.6. This software is capable of classifying satellite image by both supervised and unsupervised approaches. In this study, we opted for both supervised and unsupervised classification separately on two different remotely sensed data. An unsupervised classification approach is based on ISODATA algorithm, in which the pixels were grouped into cluster. One hundred twenty spectral classes with 12 iteration and 95% convergence values were selected to perform unsupervised classification. After classification we recoded the classified satellite image. The result comes in the form of eight classes in Landsat image and twelve in the IRS 1C LISS III. For better accuracy we combined the four classes (scrub with, scrub without, mining or industrial waste and plantation) of IRS 1C LISS III classified image into wasteland. For supervised classification prior knowledge of study area is required; for supervised classification training sites were made by demarcating a polygon or area of interest (AOI) for the know land cover/land use type, using signature editor tool in ERDAS 8.6 software. Later on, the basis of these signatures’ whole image was classified. Ground truth information was incorporated in the refinement of training sites selection during the final classification image analysis. The supervised classification approaches were based on the maximum likelihood classification decision rule. The error matrix of classified images showed that the overall accuracy was 88%, which was satisfactory. According to Anderson (1976), the accuracy of classified images should be more than 84% for better results.
The different types of land cover/land use in the classified image (Figure 2 a 2b and 2c).
Results and Discussion: LULC cover change analysis
Comparison of land use and land cover derived from images of 1989 & 2000. After classification, the comparison of different land use and land cover showed either decrease or increase in area. A decrease in the river area by ~ 3.7 % was observed. This decrease was due to increase in the area of settlement. A phenomenal increase was observed in the number of settlements, as high as > 330 % because of increase in population and also improved standards of life. There was a decrease in the cropland area by 5.64 % and an increase of fallow land by 28.7 %. The cropland area decreased because of increase in population and increase in the area of fallow land, suggesting less rainfall. Soil fertility changed due to excessive use of fertilizers. Salt affected land increased by 46 %, owing to Green Revolution in Punjab. Increase of salt affected land was due to the enormous use of groundwater, fertilizers and pesticides. The area occupied by water bodies showed a decrease of around 16.67 %.
Comparison of land use and land cover derived from images of 2000 & 2005. The area under the river in this case showed a negative trend as well, a sharp decrease of about 14 % of the total study area. The data signals an increase of settlement around 45 % of total study area. The cropland showed a decrease by 16.91 % of the study area. The dense forest cover showed a decrease by 3 % of its total area. There was a significant decrease (> 75 %) of area of water bodies in 2005, as compared to that in the year 2000.
Conclusion derived from land use and land cover study from 1989 to 2005. There was a continuous decrease in the area under river (17 %), dense forest (22 %), cropland (31 %) and a significant increase in the area under settlement (534 %). The main reason behind the change in land use and land cover pattern may be the increase in population size and per capita requirement of natural resources. The land with scrub increased considerably (214 %) over the period, while land without scrub registered a decrease of 63 %. The total increase in salt affected land was around 94 %. The area under water bodies as well as under seasonal streams displayed a negative trend, with the total decline being 81 % and 36.5 %.
Table 1 Changes in respective classes in years 2000 & 2005 with respect to 1989
The study demonstrated that there was a great decline in croplands and forests due to urbanisation and other alternative LULC. Data acquired from the average rates of all the different types of land cover/land use transitions and associated changes in soil organic carbon density were pooled for the total study region. The need of more agricultural land was required due to rapid increase in population. This need had mostly negative impact on the natural resources. So there should be scientific and traditional methods required for the sustainability of the environment.
Conclusion
In this study, the time series Landsat and IRS 1C LISS III data were used to monitor changes in LCLU in the Shiwalik hills of Rupnagar, Punjab, India where earlier there was Ropar lake with its related wetlands completely drained. An analysis of the nature and rate of land use change and its associated impact on natural resources is essential for a proper understanding of the present environmental problems at local and global scale. A prerequisite for an effective conservation strategy is a continuous and consistent monitoring programme on the state and spatial extent of groundwater with a cost efficient way. This study revealed that regular monitoring with the help of remote sensing may serve as a very critical tool to assess the magnitude and rate of local environmental changes and to quantify interactions among local land use and land cover changes, ecosystem emissions of CO2 and global climate change. The prospects for minimising or reducing environmental degradation, stabilising atmospheric CO2 concentrations and restoring damaged ecosystem for sustaining ecological goods and services. Use of the remotely sensed data for monitoring purposes requires periodic updates of two basic sources of information on LULC data and indirect and direct information about the quality and quantity of groundwater resources. These data represent useful information for resource managers to support their efforts to conserve and manage the natural resources and to advice the local community on the changing status.
Acknowledgement
This research is the part of M. Phil dissertation of the author, at School of Environmental Sciences, Jawaharlal Nehru University New Delhi, India.
References
- Patel NK, Patnaik C, Dutta S, Shekh AM and Dave AJ (2001) Study of crop growth parameters using Airborne Imaging Spectrometer data. Int J Remote Sensing 22(12):2401-2411
- Wali M., K., Evrendilek F., West T., Watts S., Pant D., Gibbs H., McClead B., Assessing terrestrial ecosystem sustainability: usefulness of regional carbon and nitrogen models. Nature & Resources, 1999, 35(4), 20-33
- Evrendilek F., Ertekin C., Agricultural sustainability in Turkey: integrating food, environmental and energy securities. Land Degradation & Development, 2002, 13(1), 61-67
- Kalnay E., Cai M., Impact of urbanization and land use change on climate. Nature 2003, 423, 528-531
- Houghton R., A., Revised estimates of the annual net flux of carbon to the atmosphere from changes in land use and land management 1850-2000. Tellus, 2003, 55B, 378-390
- Lal R., Soil erosion impact on agronomic productivity and environment quality. Critical review. Plant Science, 1998, 17, 319-464
- Pimentel D., Harvey C., Resosudarmo P., Sinclair K., Kurz D., McNair M., Crist S., Shpritz,L., Fitton L., Saffouri R., Blair R., Environmental and economic costs of soil erosion and conservation benefits. Science, 1995, 2671117–2671123
- Lal R., Soil carbon sequestration impacts on global climate change and food security. Science, 2004, 304, 1623-1627
- Jensen J., R., Introductory Digital Image Processing: a Remote Sensing Perspective, Second ed, Prentice Hall, Saddle River, NJ, 1996, 316
- Gautam N., C., Narayn L, .R., A., Landsat MSS data for land use and land cover inventory and mapping. Journal of Indian Society of Remote Sensing, 1983, 11(3) pp. 15-27.
- Gautam N., C., Channianh C., H., Land use and land cover Mapping and change detection in Tripura using satellite Landsat data. Journal of Indian Society of Remote Sensing, 1985, 6 (3 and 4), 517-528
- Alexander R., H., Milazzo V., A., Urban and regional land use analysis and census cities experiment package. MO. Progress Report E 74-1 0252, NASA-CR 1365661, 1973
- Mukherjee S., Change in groundwater environment with land use pattern in a part of south Delhi. A Remote Sensing approach. Journal Aisa-Pacific Remote sensing and GIS ,1998, 9 (2), 9-14
- Mukherjee S., A text book of environmental remote sensing. Macmillan, India Ltd. 2004, 34
- King C., Delpont G., Spatial assessment of erosion: contribution of remote sensing, a review. Remote Sensing Reviews, 1993, 7, 223–232
- Siakeu J., Oguchi T., Soil erosion analysis and modelling: a review. Transactions of the Japanese Geomorphological Union, 2000, 21 (4), 413–429
- Lunetta, R., S., Project formulation and analysis approaches. In: Lunetta, R.S., Elvidge, C.D. (Eds.), Remote sensing change detection: environmental monitoring methods and applications, Ann Arbor Press, Chelsea, MI, 1998, 318
- Muchoney D., M., Haack B., Change detection for monitoring forest defoliation. Photogrammetric Engineering and Remote Sensing, 1994, 60, 1243–1251
- Singh A., Digital change detection techniques using remotely sensed data. International Journal of Remote Sensing, 1989, 10, 989–1003
- Chen D., Stow D., A., Tucker L., Daeschner S., Detecting and enumerating new building structures utilizing very-high resolution image data and image processing. Geocarto International, 2001, 16, 69–82
- Abuelgasim A., A., Ross W., Gopal S., Woodcock C., E., Change detection using adaptive fuzzy neural networks: environmental damage assessment after the Gulf War. Remote Sensing of Environment, 1999, 70, 208–223
- Coppin P., R., Bauer M., E., Digital change detection in forest ecosystems with remote sensing imagery. Remote Sensing Reviews, 1996, 13, 207–234
- Lunetta R.,S., Elvidge C., D., Remote Sensing Change Detection: Environmental Monitoring Methods and Applications, Ann Arbor Press, Chelsea, MI, 1998, 318
- Stow D., A., Reducing mis-registration effects for pixel-level analysis of land-cover change. International Journal of Remote Sensing, 1999, 20, 2477–2483.
- Kardoulas N., G., Bird A., C., Lawan A., I., Geometric correction of SPOT and Landsat imagery: a comparison of map- and GPS-derived control points. Photogrammetric Engineering and Remote Sensing, 1996, 62, 1173–1177
- Townshend J., R., G., Justice C., O., Gurney C., McManus J., The effect of mis-registration on the detection of vegetation change. IEEE Transactions on Geosciences and Remote Sensing, 1992, 30, 1054–1060
- Schowengerdt R., A., Remote Sensing: Models and Methods for Image Processing, Second ed, Academic Press, San Diego, 1997, 522
- Anderson J., R., Hardy E., E., A Land Use And Land Cover Classification System For Use With Remote Sensor Data, U.S. Geological Survey (1976)
- Jaisawal RK, Saxena R and Mukherjee S (1999) Applications of Remote sensing technology for Landuse/Landcover change analysis. Journal of Indian Society Remote Sensing. 27(2): 123–128
