DETECTION LAND COVER CHANGES OF THE BAQUBA CITY FOR THE PERIOD 2014-2019 USING SPECTRAL INDICES

805  DETECTION LAND COVER CHANGES OF THE BAQUBA CITY FOR THE PERIOD 2014-2019 USING SPECTRAL INDICES A.B. Khalaf A. I. J. Al-Jibouri Lecturer Lecturer Dep. Soil and Water Resources. Dep. Desrtficition. Coll. of Agriculture. University of Diyala. Coll. of Agriculture. University of Baghdad Dr.ahmedbahjat@agriculture.uodiyala.edu.iq Afakibrahem1963@gmail.com ABSTRACT This study was conducted on the Land coverings of the city of Baquba and its outskirts in Diyala province, central Iraq, between latitudes 44o 42ʹ 31.78ʺ ــ 44o33ʹ 14.99ʺ and 33o41ʹ 46.66ʺ ــ 33o 48ʹ 23.18ʺ an area of 180,835 km 2 . In order to classify the earth covers, it was relied on the field survey to determine the grounding points. Also used two satellite data from Landsat 8, the first one on 23/3/2014, the second on 21/3/2019, and the production of Normalized Difference Vegetation Index (NDVI), Normalized Difference Water Index (NDWI) and Normalized Differences Builtup the Index (NDBI) maps. The results of the survey was showed five varieties are vegetation cover, agricultural land, water, buildings and barren land. They were identified and compared with the 75 land control points, The accuracy of the classification was calculated using Kappa It was 89% , and purely concluded that the use of manuals NDVI, NDWI and NDBI was useful for classifying Land coverings and detecting changes as they are considered an easy and fast method.


INTRODUCTION
The use of remote sensing techniques in the classification and detection of change in Earthbased, coatings on electromagnetic radiation through satellite images is very effective for obtaining a better understanding of the Earth's environment (18,26). Remote sensing data is one of the most important sources of available ground information, allow systematic and costeffective monitoring, and monitoring of the Earth and its resources, comparisons among different time periods and the production of maps. (1, 12). Changes in natural resources as shown in land cover and land-use change (LCLUC) are one of the most important components of global environmental change (20,27). Rapid growth in population and industrialization has led to increased exploitation of natural resources to meet the basic human societal needs (17), which requires continued monitoring and development of their management and investment programs through remote sensing sensors and geographic information systems. (21). Spectral indics were obtained from satellite images is a very simple and effective algorithm for quantitative and qualitative assessment of earth cover (6). In a study of the researcher Daniela et al. (4) using spectral indics, especially the NDVI (Normalized Difference Vegetation Index) and the NDWI(Normalized Difference Water Index) in detecting change in the earth cover in Donana, Spain for the years 2009-2010. The study found that the use of spectral indics was good Detection of change of ground cover for the study area. Kusay and Muntaha (8) found that the best indics for land degradation was the indics of modified soil cover (MSAVI) as well as (NDVI), which showed the largest percentage change at area (Husseini) irrigation project at Karbala governorate between 1989 and 2011. The researcher Manikandan (11) Use the Water Ratio Index (WRI), (NDWI) and Modified Normalized Difference Water Index (MNDWI), along supervised classification. The study was showed that the lakes of Chelekleka in India had a decrease of 1.309 km2 by 79% between 2000-2017 and NDWI was the best guide for quantitative expression Change. Either Taghreed (24) has used indics Difference Vegetation-Index (DVI), Perpendicular Vegetation Index (PVI) and Weighted Difference Vegetation Index (WDI) for detection and control of vegetation cover between 2001-2002 at Al Fit'ha north of Salah al-Din Governorate. The researcher found that the indics is better than the other indics used in the study. The objective of this study was to adopt the modern scientific methods based on the required software in the preparation of the required information to classify the Land coverings and detect the change in these covers at the city of Baquba in a sophisticated scientific method by creating land classification maps and showing the various practical applications of remote sensing and geographic information systems.

Study area
The study area was determined by field visits using the Global Positioning System (GPS).. In addition to using Google Earth, The administrative boundaries of Diyala Governorate are limited to 44º 42ʹ 31.78ʺ ‫ــ‬ 44º33ʹ 14.99ʺ and 33º41ʹ 46.66ʺ ‫ــ‬ 33º 48ʹ 23.18ʺ an area of 180,835 km 2 .

Satellite image
In this study, two satellite images were used for the satellite Landsat 8, the OLI_TIRS sensor (Path 168 and Row 37), the first was taken on 23/3/2014 and the second was taken on 21/3/2019 as in Figure 1.

Processing of satellite images
In order to obtain a spatial resolution of 15m, Landsat 8 was integrated or collected by using the Erdas Imagine V. 14 program. After that, the study area was cut. Using ArcMap 10.3.1, as in Figure 2. vegetation, compared to vegetation-free areas. The red colo absorbs heavily and reflects the nearby infrared radiation. This occurs by the chlorophyll found in the green leaves. Therefore, areas with dense vegetation cover their spectral properties in the infrared and infrared ranges, Areas with less dense or plantfree vegetation (7). The NDVI index is calculated based on the difference in the amount of reflected radiation in the nearby red and infrared channels divided by the sum of the reflection in the two channels. The value of the NDVI index is between (-1 and + 1), the value of which is close to 1 (0.

Normalized Differences Water Index (NDWI)
Higher values of NDWI indicate that water is larger than other covers And are shown with positive values. Table 1 show the values of the indics used in the study and for the periods 2014 and 2019. Table 1  When comparing Table 1 with Table 2, it is reval a decrease in the area of vegetation 2.022 km 2 (2.19%). This corresponds to the higher value of NDVI to 2014 of 0.5729. This value decreased in 2019 to 0.5682. The higher values of this index reflect vegetation, Or orchards, we deduce from this usefulness of the NDVI guide in controlling the change in vegetation over other species and this is consistent with the findings of many researchers (9, 15).   Table 1 and 4, shows an increases in the value and area of buildings respectively between 2014 and 2019. The importance of the NDBI indices in the study of change in Land coverings is indicated by many researchers (10, 19). Using Excel we were able to represent the values of the indices used in the study in the form of diagrams as reveal in Figure 9. In order to know the areas occupied by the earth cover for the study area more accurately and compare them with the maps of the spectral indices used in the study we conducted the supervised classification of the Landsat 8 satellites for the 2014 and 2019 periods and using the Erdas Imagine program. The areas of the land varieties resulting from the classification were then calculated and as in Table 5.  The results in Table 5 shows a deterioration of the vegetation cover by 2,889 km 2 and 7.41%, followed by an increase in the area of buildings, water and agricultural lands by 2.540 km 2 , 2.726 km 2 and 0.384 km 2 respectively, As for the barren land, the percentage of decline was 5.77% and an area of 2.761 km 2 . When comparing the classification with the classification maps of the indices used in the study, the decrease in the area of the vegetation cover, according to the classification of NDVI classified by 2.022 km 2 is an approach to the result of the supervised classification (2,889 km2), and there is an increase in the water area and according to the NDWI classification by (2.783) An approach to the area obtained from the vector classification (2.726 km 2 ). The NDBI map showed a decrease in the area of buildings in both cases (11).

Rating accuracy
The process of assessing the accuracy of the classification of the elements of the different satellite image particularly important in the classification of vegetation covers and land classification. Through this accuracy, we can determine the compatibility of the classification with these covers, and the possibility of relying on the map prepared and used in the future. Therefore, the stratified random sampling method was used to evaluate the classification accuracy of the vegetation cover in Baquba by taking 75 ground control points to determine this accuracy. Using Kappa to measure precision, this measure measures the degree of difference between the land-based control points taken and the changes that have been identified in the classification map of the same site and compared them (13). This is a comparative measure. Many researchers have pointed out that the value of Kappa, which is greater than 80%, is a good and appropriate classification and is recommended, whereas this value is limited to 40-80% and a medium rating of the image is shown. 40% show that this classification is poor and unreliable. The classification of 2019 was higher than the 2009 classification and was respectively (0.89, 0.88) as the overall average of the classification and this is considered good as shown in Table (6). Through this scale Kappa showed that there was a higher classification of (water, vegetation and agricultural land) for 2019 while the rest of the varieties were well classified. While the value of this measure for 2014 was also high for good by supervised classification. Table 6