Land Use and Land Cover Change Detection for Dehradun City (2004-2009) Dr
Total Page:16
File Type:pdf, Size:1020Kb
International Journal of Research p-ISSN: 2348-6848 e-ISSN: 2348-795X Available at https://edupediapublications.org/journals Volume 03 Issue 18 December 2016 Land Use and Land Cover Change Detection For Dehradun City (2004-2009) Dr. Asha Sharma Asstt. Prof. In Geography Mata Sundri Khalsa Girls College, Nissing longitude . The city is sorrounded by Introduction river Song in the east , river Tons on the According to the UDPFI guide lines , an west , Himalayan Shivalik ranges on the urban area is characterized by North and Sal forests in the south . It is population sizes greater thatn 5,000 the largest city in the hill state and well and at least 75% of the resident connected by rail and road transport . population dependent on non- Dehradun city alongwith its contiguous agricultural/primary activities for income outgrowths from an generation. Supporting populatons and "Urban Agglomeration" consisting of economic activites requires the Dehradun municipal area , Forest occupancy of land or the utilization of Research resources linked to land . Hence , Institue, Dehradun Cantonment , growth in population or tis redistribution Clement town Cantonment , Adhoiwala and growth/shifts in economic activity outgrowth and Raipur town. manifests as changes in land use and / Physiography or land cover. Studying land use / land Dehradun City is located on a gentle cover changes can therefore help undulating plateau at an average understand the trends and predict altitude settlement development patterns in any of 640 m above mean sea level. The region which in turn can guide spatial lowest altitude is 600 m. in the planning and policy making. southern Study Area part , whereas highest altitude is 1000 Dehradun city is situated in the m on the northern part. The northern southern part of Dehradun District slope within the newly formed Uttarakhand (south facing) has a gentle gradient of state . Dehradun is the interim capital of about 8 o which mostle comprises of Uttarakand , situated in the Dun Vally at coarse boulders, pebbles, gravels 300 19' N latitude and 78 o 02'E derived mailnly from Pre-Tertiary and Available online: https://edupediapublications.org/journals/index.php/IJR/ P a g e | 1637 International Journal of Research p-ISSN: 2348-6848 e-ISSN: 2348-795X Available at https://edupediapublications.org/journals Volume 03 Issue 18 December 2016 Shiwalik rocks with top soil and 1. To detect urban land use-land silt/snad. The north facing southern cover change for Dehradun city slops is steeped in gradient of about 10 during the period 2004-2009 o formed mainly of reworked Upper Methodology Shiwalik boulders and gravels with Data Procurement sand and silt.The drainage of the city is borne by two rivers namely:Bindal and Image of Map Rectification Rispana River . The direction of flow of Image to Image Rectification these Normalisation of Image I seasonal streams and nalas in the Extraction of Study Area eastern part is north to south and western part it is north to southwest Classification .The whole area is heavily dissected by Change detection a number of seasonal streams and The softwares that were employed for nalas , which are locally known as " the project were ERDAS Imagine and Khalas". Dense patches of forests exist Arc GIS 9.3 along the north and in the outer limit of Data Procurement regulated areas. The details of data used on the project Project Objectives are as follows SI No. Data Type Source 1. Toposheet 531/3 Survey of India 2. Setellite image-LISS 4 ,2004 for Dehradun , HRS , Dehradun resolution : 5.8m 3. Setellite image-LISS 4 ,2009 for Dehradun , HRS, Dehradun resolution : 5.8m 4. Guide Map-Dehradun HRS , Dehradun Available online: https://edupediapublications.org/journals/index.php/IJR/ P a g e | 1638 International Journal of Research p-ISSN: 2348-6848 e-ISSN: 2348-795X Available at https://edupediapublications.org/journals Volume 03 Issue 18 December 2016 Image Rectification and Registration Visual interpretation of the images The toposheet data was stored as an helped indentify the following major land image file in TIFF format . Using use categories for the study subsets. ArcGIS, the toposheet image was 1. Dense Forest ractified to the user defined co- 2. Open Forest ordiantes UTMWGS84 using 16 equi 3. Agriculture Cropland • spaced control points as input .The rectified toposheet was imported to 4. Agriculture Fallow ERDAS Imagine. One of the satellite 5. Plantation images was tehn rectified to the same 6. Land with or without scrub user defined co-ordinate system using 7. Built UP the rectified toposheet as reference. In order to ensure good alignment of 8. Water Body satellite images during the overlay 9. Open/vacant land operations for change detection, image Unsupervised Classification to image registration was done between the un-rectified and the rectified satellite The subset images of 2004 and 2009 images. were classified separately using ERDAS Imagine. Thus all images , toposheet and setellite images of 2004 & 2009 were In the classifier Tab of the Erdas brought to the same co-ordinate Imagine window , select Unsupervised system. The images were checked for Classification . alignment using swipe function in the A window opens for Unsupervised utility tool in the viewer window of Classification (isodata). The subset ERDAS Imagine. image to be classified is given as input. Classification of Image Output files are defined for cluster layer and signature set. These are stored as With the intent of understanding the image and signature files respectively. limitations and possibilities in In color scheme options, choose application, the images were classified approximate true color. The number of by the both the supervised and categories to be recognized is defined. unsupervised classification techniques. The greater the number of categories generated , the more accurate the Available online: https://edupediapublications.org/journals/index.php/IJR/ P a g e | 1639 International Journal of Research p-ISSN: 2348-6848 e-ISSN: 2348-795X Available at https://edupediapublications.org/journals Volume 03 Issue 18 December 2016 classification will be . However the In supervised classification, training cases where the class variations per samples of reflectance ranges / DN unit area are high , greater number of values associated to different land use categories may reduce the categories are defined by the user interpretation clarity . Also define the before the classification is performed by number of iterations and the the software. convergence threhol limits . The The subset images of 2004 and 2009 process output is generated when either were classified separately using of these criteria is met . The specified ERDAS Imagine. maximum number of iterations was 30 Samples of signatures denoting the and the convergence threshold was various land use categories are fed into specified at 0.98.The generated outpur the' Signature Editor in the Classifier tab resembles the reference satellite image of the ERDAS Imagine window. subset since the colour scheme option Samples of the same land use category chosedn was approximate tru colour. In are merged to form a representative case the colour scheme option chosen signature class. Each of such was preudo colour , the generated representative signature classes are outpur layer would appear in shades of named appropriately. In cases where grey . Image interpreation is easier with ground verifications are required , the approximate true colour option. Geo ground truthing is done by collecting linking the satellite image subset with Ground Control Points the generated output layer helps identify ( GCPs) to ensure accurate classes more accurately. For the representation of land use . software generated classes , type and Once the signatures are fed into the colour definition is done using the signature editor for all classes, the keyboard after clicking on well defined signature editor file was saved. This land use class samples on the signature editor file was given as input generated output layer. to generate the supervised Supervised Classification classification layer Available online: https://edupediapublications.org/journals/index.php/IJR/ P a g e | 1640 International Journal of Research p-ISSN: 2348-6848 e-ISSN: 2348-795X Available at https://edupediapublications.org/journals Volume 03 Issue 18 December 2016 Accuracy Assessment considerable mismatch between the The following procedure was followed to class value and the reference value, the perform the accuracy assessment of reference value code was revised to an each classified image. appropriate value after checking with In the Classifier Tab of the ERDAS the reference value , the reference Imagine windows , Accuracy value code was revised to an Assessment option was chosen . In the appropriate value after checking with Accuracy Assessment window that the reference subset image and opened , the file to be opened in the classified image. The Accuracy viewer was selected . The Viewer was Assessment Report was then generated selected using Select Viewer option in . Classification accuracy of the range View tab. In the Edit tab , the option to 80% and above is considered Create / Add Random Points was acceptable. opened . A value equal to ten times the The accuracy assessment reports land use classes to be indentified was generated for supervised classification input as the number of random points. of 2004 and 2009 subsets were 64% Geo linking the classified image with the and 68% respectively. reference subset image helped