• Drivers Of Land Use Change Detection In Texas
• Drivers Of Land Use Change Detection Pdf

. 1.4k Downloads. Abstract Urbanization is the primary driver of Land Use/Land Cover (LULC) changes throughout the world. It is arguably the most dramatic and prevalent form of irreversible land transformation. In a fast growing city like Delhi, land use changes are tremendous. Therefore, it is imperative to analyze the driving forces of such change. Along with Delhi, the South West District of Delhi has been chosen for a comparative study of LULC change from 1977 to 2014.

Landsat and Indian Remote Sensing (IRS) satellite scenes were used to perform both supervised and unsupervised classification and an overall accuracy of over 90% was achieved for all the years. In Delhi, net percent change from 1977 to 2014 was found to be +30.61% for built-up area, −22.75% for cultivated area, −5.31% for dense forest, −2.76% for wasteland and +2.41% for road/rail network.

No major net percent change was seen in open forest, scrubs/degraded forest, plantations and river/waterbody. The LULC results provide evidence of relationship between built-up area, agricultural land and wasteland over the past four decades. The effect of economic reforms of 1991 has manifested itself as a change in LULC. Overall it illustrates a characteristic picture of LULC change and its dynamics. Due to increasing human intervention on the environment, most landscapes on the Earth’s surface have been altered in some ways or the other.

As a result there is tremendous pressure on the land environment and its components. Land use describes how a parcel of land is used such as agriculture, residences or industry, whereas land cover describes the materials such as vegetation, rocks or buildings that are present on the surface (Lillesand et al. Land Use/Land Cover (LULC) studies have become key components for managing natural resources and understanding various impacts of human activities on the environment.

Automated logic web ctrl login. It’s everything they need to keep occupants comfortable, manage energy conservation measures, identify key operational problems, and validate the results. It allows building operators to optimize and manage all of their building systems - including HVAC, lighting, fire, elevators, and security – all within a single HVAC controls platform.

One of the major causes of LULC change is the rapid pace of urbanization. ‘Urbanized societies’ in which large proportion of people live in cities developed only in the 19th and 20th century. The growth in the urban to rural ratio is made possible by the movement of people from rural habitats to urban ones or a general progression of the rural area to an urban center because of economic growth and development (Davis ). The world’s urban population has multiplied more than ten times in the past century; from 224 million in 1900 to 2.9 billion in 1999 (Alberti ). According to United Nations estimates, the population living in urban areas exceeded 50% of the world total in 2006 and will approach 60% in 2020.

While the world’s urban population is expected to increase by almost 2 billion over the next 30 years, the world’s rural population is actually expected to decline slightly falling from 3.3 billion in 2003 to 3.2 billion in 2030. Thus, all future population growth for the foreseeable future is expected to be absorbed in urban areas. Most, if not virtually all of this growth, is taking place in the developing countries (Shahraki et al. Urbanization therefore, is arguably an irreversible and in most cases a permanent change that any return to rurality seems unlikely. It is one of the major factors leading to a change in the LULC of a region and has led to the development of a large number of cities (Cohen ).

Another great factor giving rise to urbanization is the rather late and as yet very incomplete industrialization of agriculture (Davis ). According to Davis, as farming becomes highly mechanized and rationalized fewer people are needed on the land. It implies that on an average the more urbanized a country, the lesser is it’s rural to urban ratio. The same picture can be seen in today’s world. The present trend of the urbanization in developing countries is especially due to rural–urban migration, the geographic expansion of urban areas through annexations, and the transformation and reclassification of rural villages into small urban settlements. The expansion of the metropolitan periphery is mainly caused both by the arrival of new migrants and by the sub-urbanization of the middle class out of the central city.

The relative importance of each of these various causes of urbanization and suburbanization varies both within and between regions and countries. As has been observed in the rest of the world, India had similar impacts of urbanization and LULC change.

The independence of the country gave further impetus to the urbanization in Indian cities such as Delhi, Mumbai, Kolkata and Chennai (Delhi Census Handbook ). Compared to western urbanization, which followed a gradual shift in economic base from agriculture to industry and then to tertiary sector propelled economic development, the economic development in Indian cities is being triggered by service sector growth. The economic liberalization policy of 1991 opened up Indian economy to the international market, which saw incoming of large share of foreign direct investment (FDI) in metropolitan cities. Delhi region received maximum share of FDI compared to other regions of the country. Along with this allowance of 100% FDI in real estate and infrastructure by Indian government has made Delhi susceptible to rapid urban growth (Namperumal et al.

Consequent reduction in agricultural and allied activities has led to reduced output from the primary sector to Delhi’s economy. This predominant shift from agriculture to service has created abundant job opportunities attracting people from different states in search of employment. Cultural model of elitism is yet another factor that attracts people to the city. Further, population has been increasing at an exponential rate in the national capital. Delhi’s population has increased from a meager 405,800 in 1901 to a staggering 16,753,200 in 2011 (Census of India ), making it one of the largest growing cities in the world. Statistics show that 2.22 million immigrants entered Delhi from 1991 to 2001, which is substantially higher than the 1.64 million who had come in from 1981 to 1991 (Delhi Human Development Report ).

Delhi, unlike many cities, caters not only to its residents, but to an unusually large number of outsiders as well. The city like many growing urban centers attracts a large inflow of migrants from all over India, especially from Bihar, Haryana, Rajasthan, and Uttar Pradesh. An interesting yet unusual feature of Delhi is the large floating population of millions of visitors every year.

These short-term visitors are drawn to the city for a variety of reasons including for work and business, for accessing healthcare, for sightseeing and studying. With such dramatic changes in economic policies, population growth and influx along with development of the national capital, it begs the question of whether these can be seen in concurrence with LULC change or not. While the ground survey method of data collection is usually the traditional method of monitoring land use and urban growth (Olorunfemi ), classic land-cover change detection techniques are based on the comparison of sequential land-cover maps derived from remote sensing (Yang and Lo ). As a result of the organization problems and time, ground surveys are not possible to undertake in quick succession; hence they do not yield the needed time series data. Meanwhile, this drawback can be easily resolved with the use of remote sensing data. Therefore, the role of techniques such as remote sensing, aerial photography and image processing becomes important.

Remote sensing and image processing have been used to document urban growth and dynamics in hundreds of areas throughout the world over the last decades (Miller and Small; Sudhira et al.; Fox and Vogler; Jat et al.; Zeng et al.; Lasaponara and Lanorte ). Timely and accurate information on existing LULC pattern, its distribution and changes over time is a pre-requisite for planning, utilization and formulation of policies and programs for making any developmental plan.

More often than not in developing countries, government data proves to be insufficient, inaccurate, obsolete or simply non-existent (Rahman et al.; Pandit et al. In such a case where only a few authentic data sources are available in hand, use of satellite imageries proves to be the best solution. A few satellites providing multi-temporal, multi-spectral and multi-resolution data are Landsat, IRS, Quickbird, MODIS, SPOT and IKONOS etc.

Therefore the present study is focused at preparing a multi-temporal GIS database for LULC change and to assess the spatial and temporal LULC changes in Delhi between 1977 and 2014 using remote sensing and GIS techniques. Primary aim of the study is to quantify the changes in various LULC classes over the period of four decades in the study area and to analyze the driving forces of such change. The following section briefly describes the region of the study. Section gives a detailed account of the datasets used and the methodology. The LULC results and their interpretations along with accuracy assessments for all the years of the study are presented in Sect. Section provides the final conclusions of the study.

The study area (Fig. ) covers an area of 1483 km 2, which falls under Delhi Metropolitan city as per Census of India, 1991. Delhi is the national capital of India and is located in the coordinates of 76.84°E, 28.41°N, by 77.35°E, 28.88°N. According to Census of India ( ), Delhi is comprised of nine administrative districts—North West, North, North East, East, New Delhi, Central, West, South West and South.

Besides Delhi is adjoined by couple of large districts which form an integral part of the National Capital Region (NCR). Some of the large satellite cities adjoining Delhi are Gurgaon, Noida, Faridabad and Ghaziabad. In the larger geographic context Delhi is situated between the Himalayas in the north and Aravalis in the south and the river Yamuna in the eastern part.

The elevation of the city ranges between 213 and 290 m. The area is characterized by hot and dry summers and fairly cold winters with summer time high temperatures that can reach 45 °C. Rainfall is dominated by monsoonal weather pattern, with maximum rainfall occurring from June to September. Fig. 1 Maps showing ( a) the study area enclosed in NCR region and ( b) the nine administrative districts of Delhi In addition to Delhi, LULC change detection of South West district of Delhi was studied separately because of the fact that over the last two decades the district has been experiencing tremendous growth in built-up areas including new colonies, roads, metro rails.

The South West District of National Capital Territory (NCT) of Delhi is situated in the South West part of Delhi. It is located in the coordinates of 76.84°E, 28.48°N by 77.22°E, 28.67°N. The district occupies an area of approximately 420 sq. (Census of India ).

The district is divided into three administrative subdivisions—Delhi Cantonment Sub-Division, Vasant Vihar Sub-Division and Najafgarh Sub-Division. Satellite imageries of the study area were acquired for four distinct years viz.

1977, 1993, 2006 and 2014 (Table ) from either Landsat or IRS platforms. Alongside raw satellite data, ancillary data comprising Survey of India toposheet at 1:50,000 scale and Master Plan of Delhi was gathered. Ancillary data helps to provide a basic knowledge of the area of interest and various permanent features like road/rail network, rivers etc. Were transferred to prepare a base map. Administrative state boundary map of NCT of Delhi was prepared and brought to UTM projection in zone 43North. The Delhi state boundary and boundary of South West district of Delhi were used later on to extract the area of interest (AOI).

Image preprocessing methods included initial processing of raw image data to correct for geometric distortions, to calibrate the data radiometrically, and to eliminate the noise present in the data. Since 1993 image was in polyconic projection it was brought to the same projection as the other imageries i.e. Preparation of thematic maps from digital satellite data of the 4 years were carried out on ERDAS Imagine ver. 9.1 and ArcGIS ver. Bands NIR, Red and Green of each of the imageries were used to create the False Color Composite (FCC) maps.

Figures and show the FCC’s of study area for the years 1977, 1993, 2006 and 2014. Image enhancement techniques of brightness and contrast manipulation were used for better visual interpretation. Unsupervised classification scheme with ten subsequent iterations were performed on all the generated FCC’s. Spatial and spectral pattern recognition, which includes use of elements like tone, texture, shape and location of pixel groups in the image were adopted for identification of LULC classes. Nine different Land Use classes were prepared—dense forest, open forest, scrubs/degraded forest, plantations, cultivable area, built-up area, road/rail network, river/waterbody and wasteland. The classification of forest was done on the basis of canopy cover. It is often seen that certain features like dense forest is misinterpreted as water.

Therefore for a correct final output these and any other misinterpreted patches amongst all the classes need to be recoded to the correct class. In addition linear features like road/rail network have also been mapped using recoding techniques giving rise to a non-grainy class output. Thus, supervised classification was carried out on the preliminary unsupervised classification output to produce the final LULC output class for the 4 years. This kind of hybrid classification scheme though time consuming produces highly refined and accurate class outputs as compared to either unsupervised or supervised classification alone. Aerial photographs and other photographs were documented as a part of the field survey and ground checking. GPS measurements were taken alongside to validate land use form as seen in photographs with the form as obtained through analysis of satellite imageries.

Fig. 4 False color composites of South West District of Delhi for the years ( a) 1977, ( b) 1993, ( c) 2006 and ( d) 2014 Image post-processing The classified output is smoothed using a 3.3 pixel medium statistical filter to show the majority or the dominant class in the window pass. After carrying out smoothening procedure, the raster LULC outputs are converted to vector formats. In the next step the vector LULC classes are clipped with the AOI boundary and the final LULC map is extracted.

Subsequent accuracy assessment was done to check the overall classification accuracy rate. It has been discussed in further detail in the results section. LULC change analysis was carried out for Delhi and South West district of Delhi using the time series output thus generated. Results and discussion.

The detailed classified images of Delhi depicting nine different LULC classes for the years 1977, 1993, 2006 and 2014 are shown in Fig. In the span of approximately four decades, enormous changes in LULC pattern can be observed.

The city has expanded in terms of built-up area. A radially outward growth from the earliest urban establishments viz.

Old Delhi, Mehrauli and Shahdara can be noted over the years. Urban expansion has reached to a situation of near spatial saturation in the trans-Yamuna region, while the relatively new constructed colonies like Dwarka and Rohini are still showing signs of continual growth. Results of LULC change in Delhi are presented in Table.

Similar study on the South West District of Delhi provided a comparative analysis on the differences in LULC change in a single district in contrast to the whole city. Figure shows LULC maps for the district while the statistics are compiled in Table. Several interesting results are presented. A confusion matrix scheme is used for quantitative accuracy assessment of image classification for all the years. A stratified random sampling method is used to create 50 sample verification points for each of the nine LULC classes following Currit ( ). The verification and original classification pixels are along x-axis and y-axis respectively in the resultant matrix (Tables, ).

Errors of omission, errors of commission, consumer’s accuracy and producer’s accuracy have been calculated for each class. The classified images of 1977, 1993, 2006 and 2014 have an overall classification accuracy of 91.33, 91.33, 91.11 and 91.55% respectively. On the other hand, the Kappa Index of Agreement (KIA) of the classified images of 1977, 1993, 2006 and 2014 were calculated as 90.25, 90.25, 90.00 and 90.50% respectively.

The KIA values which incorporate errors of omission and commission shows slightly lower accuracy level than overall classification accuracy but this level of accuracy is considered to be very good (Guerschman et al. 1 2 3 4 5 6 7 8 9 Total EC CA 1 39 2 1 0 0 5 0 3 0 50 0.220 0.780 2 2 46 1 1 0 0 0 0 0 50 0.080 0.920 3 1 1 44 1 0 3 0 0 0 50 0.120 0.880 4 0 0 0 48 2 0 0 0 0 50 0.040 0.960 5 0 3 0 0 46 1 0 0 0 50 0.080 0.920 6 0 0 2 1 0 46 0 0 1 50 0.080 0.920 7 1 0 0 0 1 0 47 1 0 50 0.060 0.940 8 0 1 0 0 0 0 0 49 0 50 0.020 0.980 9 0 0 0 1 3 0 0 0 46 50 0.080 0.920 Total 43 53 48 52 52 55 47 53 47 450 EO 0.093 0.132 0.083 0.077 0.115 0.164 0.000 0.075 0.021 PA 0.907 0.868 0.917 0.923 0.885 0.836 1.000 0.925 0.979. 1 2 3 4 5 6 7 8 9 Total EC CA 1 42 2 0 1 0 0 0 5 0 50 0.160 0.840 2 1 44 1 1 3 0 0 0 0 50 0.120 0.880 3 0 1 44 1 0 1 3 0 0 50 0.120 0.880 4 0 0 1 47 0 1 0 0 1 50 0.060 0.940 5 0 3 0 1 43 0 1 0 2 50 0.140 0.860 6 0 0 0 1 0 46 3 0 0 50 0.080 0.920 7 0 0 0 0 1 1 48 0 0 50 0.040 0.960 8 0 0 0 0 0 0 0 50 0 50 0.000 1.000 9 0 0 0 0 2 1 0 0 47 50 0.060 0.940 Total 43 50 46 52 49 50 55 55 50 450 EO 0.023 0.120 0.043 0.096 0.122 0.080 0.127 0.091 0.060 PA 0.977 0.880 0.957 0.904 0.878 0.920 0.873 0.909 0.940. 1 2 3 4 5 6 7 8 9 Total EC CA 1 49 1 0 0 0 0 0 0 0 50 0.020 0.980 2 2 41 5 0 1 1 0 0 0 50 0.180 0.820 3 0 2 44 3 0 0 0 0 1 50 0.120 0.880 4 0 1 0 44 1 2 1 0 1 50 0.120 0.880 5 1 0 0 3 42 0 0 0 4 50 0.160 0.840 6 0 0 0 2 0 45 2 0 1 50 0.100 0.900 7 0 0 0 0 0 1 49 0 0 50 0.020 0.980 8 1 0 0 0 0 0 0 49 0 50 0.020 0.980 9 0 0 0 0 1 2 0 0 47 50 0.060 0.940 Total 53 45 49 52 45 51 52 49 54 450 EO 0.075 0.089 0.102 0.154 0.067 0.118 0.058 0.000 0.130 PA 0.925 0.911 0.898 0.846 0.933 0.882 0.942 1.000 0.870. 1 2 3 4 5 6 7 8 9 Total EC CA 1 43 4 1 2 0 0 0 0 0 50 0.140 0.860 2 1 46 0 2 1 0 0 0 0 50 0.080 0.920 3 0 0 42 0 0 0 7 1 0 50 0.160 0.840 4 0 4 0 43 1 0 0 2 0 50 0.140 0.860 5 3 0 0 2 45 0 0 0 0 50 0.100 0.900 6 0 0 2 0 0 48 0 0 0 50 0.040 0.960 7 0 0 0 0 0 0 50 0 0 50 0.000 1.000 8 1 0 0 0 0 0 1 48 0 50 0.040 0.960 9 0 0 0 2 0 1 0 0 47 50 0.060 0.940 Total 48 54 45 51 47 49 58 51 47 450 EO 0.104 0.148 0.067 0.157 0.043 0.020 0.138 0.059 0.000 PA 0.896 0.852 0.933 0.843 0.957 0.980 0.862 0.941 1.000. Change detection refers to a group of techniques by which significant differences in digital image values in multi-temporal images are assessed.

From-to algorithm describes LULC classes present before and after a quantitative change in digital image values whereas the other group of algorithms simply detects the presence or absence of change (Currit ). Net percent change in LULC classes from 1977 to 2014 for Delhi have been calculated and shown in Fig. Most LULC changes are complex and multidirectional. It may be possible that the net percent change in area of a certain class showed no considerable change temporally, but did so in spatial terms. Plantations is one such class that can decrease in spatial extent in one area and increase in another area, thereby showing no quantitative change.

But urban parameters like road/rail network and built-up area donot usually get demolished and simultaneously erected elsewhere. In the bigger picture, it is possible to extract valuable LULC change information from the net percent change. Fig. 7 Net percent change in LULC classes from 1977 to 2014 in ( a) Delhi and ( b) South West District of Delhi From Fig., it can be easily noted that built-up area and cultivable area are two classes showing phenomenal net percent change in area. As much as +30.61% net percent change in Delhi and +27.35% in South West District can be seen in the class of Built-up area. While for Cultivated area the net percent change stands at −22.75% for the city and −24.34% for the district.

Considerable changes can be observed in dense forest, scrubs/degraded forest, road/rail network and wasteland. Net percent change varied between −1.70 and −5.31% in case of forested area and in case of Wasteland it amounted to −2.76% for Delhi and −2.08% for South West District.

Road/Rail Network has changed to +2.41% in the city and +2.72% in the district from 1977 to 2014. The remaining classes show minimal change in terms of net percent area over time. Relationship between built-up area, cultivable area and wasteland. From the analysis of LULC data for Delhi as well as South West District of Delhi it could be gathered that there is a relationship between the three classes viz.

Built-up area, cultivable area and wasteland. Delhi’s economy has evolved from a predominantly primary sector that was based on agriculture and allied activities to tertiary sector as the opportunities for services and jobs were created. The effect of changing economy has manifested itself as a change in LULC. Table gives the comparative area of the Built-up area and the sum of the other two remaining classes.

The relationship between these three LULC classes is depicted through time plots as shown in Fig. A, b, built-up area, cultivable area and wasteland have been plotted individually over time whereas in Fig. C, d, the sum of ‘cultivable area and wasteland’ along with built-up area has been taken. In the time span of 1977–2014, a general trend of increasing built-up area can be seen in both Delhi as well as South West District of Delhi; the same cannot be inferred for the other two classes. Though if those two classes are combined then the result is far more appreciable.

Drivers Of Land Use Change Detection In Texas

It can clearly be noted from Fig. C, d that as the sum of ‘cultivated area and wasteland’ decreases, the built-up area increases. Practically, no other class has any major impact in the relationship between these three classes.

Drivers Of Land Use Change Detection Pdf

It suffices to say that virtually all urbanization taking place in the city as well as the selected district, is on the land previously occupied by cultivated and wasteland. Land for expansion came from clearing agricultural land which led to increase in wasteland. Therefore, new colonies and existing colonies then sprawled on the empty wasteland. To further strengthen the hypothesis, graphs (Fig. ) between built-up area and ‘cultivated area + wasteland’ were drawn.

It establishes a near perfect negative correlation between the two variables. The statistical analysis is tabulated in Table.