INTERNATIONAL JOURNAL OF ENVIRONMENTAL SCIENCES Volume 6, No 5, Copyright by the authors - Licensee IPA- Under Creative Commons license 3.

Transcription

1 INTERNATIONAL JOURNAL OF ENVIRONMENTAL SCIENCES Volume 6, No 5, 2016 Copyright by the authors - Licensee IPA- Under Creative Commons license 3.0 Research article ISSN Normalised difference water index based water surface area delineation and assessment Mohit Singh, S.N.Mohapatra and Monika Sharma Centre of Remote Sensing and GIS, School of Studies in Earth Science, Jiwaji University, Gwalior , India mohit_singhk@yahoo.co.in doi: /ijes.6060 ABSTRACT In the present paper, the geospatial technology has been extensively used to map and assess the changes of water surface area of Tighra dam situated near Gwalior city of Madhya Pradesh.. The change of water surfaces area has been examined for last one decade period. The LANDSAT digital satellite data from different sensors for the year 2004, 2009 and 2014 have been used for the present study. All the satellite data used are of the month June of respective years. The Normalised Difference Water Index (NDWI) was derived by using the NIR and SWIR-1 band. The NDWI along with the SWIR-2 band has been used to delineate the water surface area. In order to delineate the water surface, a threshold NDWI value has been assigned to separate it from other classes. The possible causes for the changes in the water surface area over the said years have been discussed. The climatic conditions particularly the rainfall over the study period has been studied and correlated with the fluctuation of the water surface area of the dam It has been concluded that the changes in open water features of the area is related to climate change and other societal development activities. However, it is observed that the resolution of the images is one of the constraints, and high-resolution image can provide more accurate and precise result. Keywords: Water Surface Area, Water Resources, LANDSAT, GIS, NDWI. 1. Introduction Water plays such a significant role in our day to day life, that it has been inevitable to manage and monitor water resources.. But the availability of water and its amount continuously fluctuate whole year. So, monitoring of this fluctuation of water availability over a large area is required to avail the water to us without interruption. Dams are the engineering structures, built to create a reservoir to meet the demands of water for human needs. But monitoring these resources using conventional techniques is a tough task. However, field measurements of the dam level are still overpriced and involve a lot of effort and time (Yunus and Fidelia, 2012). In this era of geospatial technology, we can estimate, monitor and map the surface water bodies occurrences and its extent with the availability of suitable satellite data. By acquiring synoptic view of the earths surface using remote sensing technologies helps to detect not only changes but also provides cheap and quick information about the status of the water in reservoir/dams and the land use/land cover (Mustafa et al., 2012; Verma et al., 2013). In the present study, the open water surface area of the Tighra reservoir has been estimated using Normalized Difference Water Index (NDWI) and multispectral LANDSAT satellite data of Received on December 2015 Published on March

2 the year 2004, 2009 and This study further helps in prioritizing the surface water body for conservation purposes. 2. Study area Tighra Dam is situated in the west of Gwalior City at N latitude and E longitudes on the Sank River. The climate of Gwalior is extremely hot and cold during summers and winters respectively (Figure 1). It experiences south-western monsoon rains in June-September with an annual rainfall of around 766 mm. Figure 1: Climate data for Gwalior from year : Source IMD. Figure 2: Location of Study Area. 635

3 Tighra reservoir forms the main source of water for Gwalior city (Figure 2). The water of the reservoir is also dependent on the Tighra-Kaketo system. The supplies from Tighra dam are supplemented by supply from another reservoir of the Kaketo dam on Narver River. Water is supplied from the dam to the city after treatment of water from Moti Jheel plant. A significant volume of this water is supplied to Gwalior city from the Tighra dam. Thus water levels in the dam are checked on a daily basis for assuring that there should be no shortfall in supplies for next coming days. 3. Methodology 3.1 Data Used In this study, LANDSAT satellite data, topographic sheet and other ancillary data such as weather and climatic condition of the study area has been used.. Any change detection analysis requires at least two data of different time period. Here we used LANDSAT sensors data (Table 1) of the month June of the year 2004, 2009 and Details characteristics of these data are discussed below. Table 1: Summarised detail of Landsat sensor data used in this work S.No. Acquisition Month Satellite Sensor No. of Bands Band Used for NDWI / Resolution (meter) 1 June 2004 Landsat 7 ETM+ 8 Band4/30m Band5/30m NIR SWIR-1 2 June 2009 Landsat 5 TM 7 Band4/30m Band5/30m NIR SWIR-1 3 June 2014 Landsat 8 OLI TIRS and 11 Band5/30m Band6/30m NIR SWIR Landsat 7 ETM+ The Landsat Enhanced Thematic Mapper Plus (ETM+) is the sensor of the Landsat 7 satellite. The images of earth have been acquired nearly continuously since July 1999, with a 16-day repeat cycle. The images of Landsat 7 are referenced to the Worldwide Reference System-2. The Scan Line Corrector (SLC) on the instrument failed in May of 2003 ( This caused all Landsat 7 data acquired after this date to have line gaps. Landsat 7 ETM+ images have eight spectral bands with a spatial resolution of 30 meters for bands 1 to 7. The panchromatic band 8 has a resolution of 15 meters. All the 7 bands collect one of the two gain settings (high or low) for increased radiometric sensitivity and dynamic range. Whereas, Band 6 collects both high and low gain for all scenes. Approximate scene size is 170 km north-south by 183 km east-west. The acquired data of

4 June-2004 of path 145 and row 042 has been used for this study. The data is affected by the SLC-off error, and it is filled by local histogram matching and focal analysis method using image processing techniques Landsat 5 TM The Landsat Thematic Mapper (TM) sensor of Landsat 4 and 5 was operated from July 1982 to May 2012 with a 16-day repeat cycle. The images were referenced to the Worldwide Reference System-1. Very few images were acquired from November 2011 to May The satellite began decommissioning activities in January Landsat 4-5 TM image data files consist of seven spectral bands. The resolution is 30 meters for bands 1 to 7. (Thermal infrared band 6 was collected at 120 meters but was resampled to 30 meters ( The approximate scene size is 170 km north-south by 183 km east-west. In this study, the has been acquired data of path 145 and row 042 of dated 12- June Landsat 8 OLI and TIRS The Operational Land Imager (OLI) and Thermal Infrared Sensor (TIRS) are instruments onboard for the Landsat 8 satellite, which was launched in February of The satellite collects images of the Earth with a 16-day repeat cycle, referenced to the Worldwide Reference System-2 ( The satellite s acquisitions are in an 8-day offset to Landsat 7. The approximate scene size is 170 km north-south by 183 km east-west. The spectral bands of the OLI sensor are similar to that of Landsat 7 ETM+ sensor which provides enhancements except two new spectral bands: a deep blue visible channel (band 1) specifically designed for water resources and coastal zone investigation, and a new infrared channel (band 9) for the detection of cirrus clouds. Two thermal bands (TIRS) capture data with a minimum of 100-meter resolution but are registered to and delivered with the OLI data product. Landsat 8 file sizes are larger than Landsat 7 data, due to new bands and improved 16-bit data product. In this study, we have acquired the data of 10-June-2014 for 145/042 (path/row). 3.2 Methods The flowchart of methodology is shown in the (Figure 3). The major steps are shown in the flowchart but there are also some intermediate steps which will be carried out during the tenure of study are not shown Geometric Correction Landsat data of three different years are georeferenced with Survey of India topographic sheet which is having scale 1:50000 and re-projected to UTM Zone 44 North considering the 16-GCPs for the image to map registration using first order polynomial transformation and nearest neighbor sampling method Filling Gap of Landsat 7 ETM+ Data Due to the failure of instrument Scan Line Corrector in Landsat 7 ETM+ sensor imagery are affected by line gap. These gaps are filled by applying linear histogram matching methodology to the filling image to adjust it based on the standard deviation and mean values 637

5 of each band. In this digital image processing techniques, LANDSAT 5 TM scene considered as the master scene for filling the gap. Figure 3: Methodology 638

6 3.2.3 Atmospheric Correction Many atmospheric correction methods are widely used for pre-processing satellite data to enhance the imagery in terms of visual interpretation. But there is dark object subtraction which is often useful when we perform band-ratios. The dark object represents the area of zero reflectance values in the image including reflection from water in NIR, shadow, scattering and electrical gain in the sensor. In this work, dark object subtraction was applied by verifying the dark object greatest in reflected band and with comprises less in the other band NDWI and Thresholding There are many band rationing technique which are proposed for use with multispectral satellite data. Some are more complex ratios involving the sums of and differences between the various spectral band, from one of them, is Normalized Difference Water Index (NDWI) has been widely used for estimating water of vegetation using remote sensing. NDWI is defined as (ρ(0.86 µm) - ρ(1.24 µm))/(ρ(0.86 µm) + ρ(1.24 µm)), where ρ represents the radiance in reflectance units(gao, 1996). Figure 4: NDWI for year (a) 2004, (b) 2009 and (c) The value of NDWI ranges from -1 to +1. NIR SWIR 1 NDWI = NIR + SWIR 1 In the resultant NDWI images has been shown in Figure 4, the pixel which is having higher water content is appears brighter than the pixel having lower water content. The NDWI value for the year 2004 ranged from to +0.15, 2009 ranged from to and

7 ranged from 0.7 to 0.5. For the delineation of surface water extent, we estimated and apply thresholding to the range of NDWI by relating the range of brightness values of SWIR-2 band of these Landsat sensors where water bodies appeared nearly black in grey scale. Subsequently, the water surface area of Tighra reservoir (Figure 5) is estimated after applying thresholding to the NDWI ranges. Figure 5: Estimated water surface area for year 2004, 2009 and Results and discussion In this study, there is a decrease in water surface area of Tighra reservoir which is estimated from the year 2004 to 2014 is sq.km. (Figure 6). Water surface area of the Tighra reservoir in the year 2004, 2009 and 2014 was 13.17, 12.94, sq.km. respectively (Table 2). The average estimated surface area of Tighra reservoir for the period of is sq.km. Figure 6: Estimated water surface area for the period of

8 Figure 7: Change in water surface extent in Tighra reservoir from Meteorological data has been correlated with the resultant change in water surface area. There are changes in the surface water extent of the Tighra reservoirs from 2004 subsequently in ten years are shown in Figure 7. During this period , we have noticed that there is a little increment in average total rainfall comparatively to a single year and also rise in temperature. Total monthly precipitation (Figure 8) of the Tighra reservoir region is considered according to the year of satellite data used in this study. Total rainfall in Tighra reservoir region for the year 2004, 2009 and 2014 was 735.6, and mm respectively which has been given in Table 2. When the minimum, maximum and average rainfalls for the five-year are considered it is observed that even after good rainfall condition 641

9 water surface extent tends to decrease in the reservoir. This increase in rainfall does not bring trend up and increase the water surface area rapidly, but gradually it takes the time to reflect the change in water surface area of the reservoir. Table 2: Yearly and periodical analysis of data for total rainfall, average temperature and estimated water surface area Year/Period Rainfall (mm.) Temperature ( C) Estimated Water Surface Area (sq.km.) Figure 8: Total monthly rainfall for year 2004, 2009 and The average rainfall for a period of ten years is mm in the Tighra reservoir region. The total rainfall for the year 2004 is mm which is lower than the average rainfall. Similarly, the total rainfall recorded in the year 2009 is mm and in the year 2014 is mm which is also lower than the average rainfall. Thus, lower rainfall in the region can be considered leading cause of the decrease in water surface area of Tighra reservoir. Annual average temperature variation for the period is shown in Figure 9. On examining the mean temperature data for the Tighra reservoir region it was C in the year 2004, C in the year 2009 and C in the year 2014 (Table 2). The average temperature of the Tighra dam region for the period was C. Except in the year 2004, the mean temperature of the year 2009 and 2014 were higher than the mean temperature for the period Thus rise in temperature causes increase in evaporation rate to decrease water surface area of the Tighra reservoir. In Figure 10 average annual temperature of each year from the period and its trend which is went down are observed during analysis. Similarly while carrying out the statistical analysis of meteorological data as a minimum, maximum and the average temperature for the period of and shown in Table 3, there is a decrease in mean temperature values are observed. The relationship between rainfall, temperature and water surface area extent are shown in Figure 642

10 10. Thus, a decrease in water surface area cannot be directly co-related to these meteorological data except for rainfall. There are other factors such as supply and demand of water for irrigation, industrial and domestic purposes causing changes in water surface area of the reservoir. Figure 9: Annual average temperature. Table 3: Five-year maximum, minimum, average of rainfall, temperature and estimated change in water surface area Period Max. Min. Avg. Rainfall (mm.) Temperature ( C) Estimated Change in Water Surface Area (sq.km.) Figure 10: Relationship between Rainfall, Temperature and Water Surface Area Extent. 643

11 5. Conclusions Monitoring the changes of water in the reservoir is important in terms of ecological and historical consideration. In this study, ten year period was investigated easily with the help of Remote Sensing and GIS technology. Landsat sensors data of the year 2004, 2009, 2014 and meteorological data was considered, and it is concluded that there is a reduced open water surface area from to 12.15sq.km in ten years from 2004 to The reason for this declined in water surface area is due to change in climatic condition. The rainfall is a major factor which is responsible for increase/decrease of water in the Tighra reservoir and also the supply and demand scenario of consumption of water for irrigation, industrial and domestic purposes. The major part of Gwalior city which mainly depend upon the supply of water from Tighra reservoir are now facing a shortage of water supply. The Municipal Corporation, Gwalior provides water on alternate days from the dam to ensure the water should be available in next coming days and hope the reservoir will be filled in next upcoming rainy season. Thus, continuous monitoring of water in the reservoir is essential in order to take necessary action. Field measurements of the dam level are still very expensive and requires a lot of effort and time. This study shows that the synoptic view of Satellite data and the use of geospatial technology help to provide information on the larger surface area of the earth without conducting a field survey. NDWI is less affected by atmospheric scattering. NDWI does not remove completely background soil reflectance effects, but it minimizes the effect of their reflectance which makes it easier to delineate and estimate the surface area of water bodies on the Earth s surface ( There are other factors such as the resolution of remote sensing data which are the main constraints for more accurate and precise work. Delineating the water surface area from satellite imagery has its own advantages over the conventional techniques. With the help of remotely sensed data, we can easily interpret the water bodies in the larger area of the earth surface. In the perspective of remote sensing and geographical information system (GIS), the spectral and spatial characteristics of these satellite data helped in wetland mapping, estimation of moisture condition of soil and vegetation with the aid of digital image processing techniques. Remote Sensing data not only helps in the study of the larger surface area of the earth but also helps in temporal change analysis with ease if there is no previous dated data is available. Thus, it can play an important role in monitoring changes in water resources and help to take necessary action 6. References 1. Gao, B. C. (1996). NDWI a normalized difference water index for remote sensing of vegetation liquid water from space. Remote sensing of environment, 58(3), pp Landsat 8 OLI (Operational Land Imager) and TIRS (Thermal Infrared Sensor). (n.d.). Retrieved January 22, 2016, from 3. Landsat Enhanced Thematic Mapper Plus (ETM). (n.d.). Retrieved October 26, 2014, from 4. Landsat Thematic Mapper (TM). (n.d.). Retrieved October 26, 2014, from Y. T., Ali, R. T., and Saleh, R. M. (2012). Monitoring and evaluating land cover change in the Duhok city, Kurdistan region-iraq, 644

12 by using remote sensing and GIS. International Journal of Engineering Inventions, 1(1), pp Mustafa, Y. T., and Noori, M. J. (2013). Satellite remote sensing and geographic information systems (GIS) to assess changes in the water level in the Duhok dam. International Journal of Water Resources and Environ Engineering, 5(6), pp Salami, Y. D., and Nnadi, F. N. (2012). Reservoir storage variations from hydrological mass balance and satellite radar altimetry. International Journal of Water Resources and Environmental Engineering, 4(6), Scaramuzza, P., Micijevic E, and Chander, G. (2004). SLC Gap-Filled Products Phase One Methodology. Methodology.pdf. 8. Verma, A., Thakur, B., Katiyar, S., Singh, D., and Rai, M. (2013). Evaluation of ground water quality in Lucknow, Uttar Pradesh using remote sensing and geographic information systems (GIS). International Journal of Water Resources and Environ Engineering, 5, pp Xu, H. (2006). Modification of normalised difference water index (NDWI) to enhance open water features in remotely sensed imagery. International Journal of Remote Sensing, 27(14), pp