Detection of Physical and Chemical Parameters Using Water Indices (NDWI, MNDWI, NDMI, WRI, and AWEI) for Al-Abbasia River in Al-Najaf Al-Ashraf Governorate Using Remote Sensing and Geographic Information System (GIS) Techniques

Abstract


Introduction
Water is an important natural resource for human survival and ecosystems, including coastal and freshwater areas (lakes, rivers, and groundwater). Over the last few decades, increased anthropogenic activity, particularly in industrial areas, has had an impact on water bodies. As a result, a river water quality monitoring program is required to promote public awareness by addressing the effects of existing and future pollution concerns [1]. Methods of field assessment of water quality indicators are often limited due to the high expense of data collection and laboratory analysis, particularly in the geographical and temporal domains [2]. An increasingly effective analysis of satellite imagery (landsat8 OLI) and geographic information systems (GIS) has developed as a significant tool for monitoring and predicting water quality parameters in rivers and streams over the last few decades [3].
The spectral index is a sum of surface reflections at many wavelengths [4]. Water parameters (PH, TDS, EC, Na, Mg, K, SO 4 , NO 3 , cl, and TSS) were collected from 10 locations along the Al-Abbasia River. In this investigation, the following indices were used: Normalized difference water index (NDWI), Modification of normalized difference water index (MNDWI), Normalized difference moisture index (NDMI), Water ratio index (WRI), and Automated water extraction index (AWEI). The wavelength spectra were extracted from the image.

Study Area
Al-Najaf Al-Ashraf city is located in southern Iraq, between latitudes (29 50' 00" -32 21' 00") north and longitudes (42° 50' 00" -45° 44' 00") east, as shown in Fig. 1. The Euphrates River, which splits into the Kufa and Al-Abbasia rivers roughly 8 kilometers before Al-Najaf Al-Ashraf province, is one of Iraq's main sources of water. The study area is in the Kufa district of Al-Najaf AL-Ashraf governorate, in the northeastern part of the governorate. It is bordered to the north-east by Babil governorate, and by Al-Qadysia governorate to the south-east and from the west by the center of Kufa district. Before the river enters the province of Al-Qadysia at the district of Al-Salahia it is known as Shatt Al-Shamiya. The study are extends astronomically between longitudes (44.22 -44.37) to the east and the latitude (31.58 -32.11) north.

Samples Collecting and Laboratory Work
Ten water samples were taken from various geographical locations along the Al-Abbasia River on dates comparable with the research area's Landsat-8 OLI dates. Table 1 shows the physical and chemical parameters for 10 sampling stations along the Al-Abbasia River that were analyzed in the laboratory.

Software Work and Satellite Image Analysis 1-Using Software
In this study geographical information system (Arc GIS) software was used and digital maps were created by (Arc Map 10.4.1).

2-Satellite Image Details
The Landsat-8 OLI satellite was used as it was timed to capture the aerial image of the study area on the same day as the samples were collected. The details of Landsat-8 satellite image were (30×30m resolution), on path 168, and row 38 and its date was 23/12/2021. The representative area of the study region was chosen from the satellite image for its clarity, ease of handling, and low data volume, as well as the ease with which the critical digital data processing activities in (Arc Map) software could be carried out.

3-Satellite Image Process
By treating reflection with special equations, processing and correcting the solar angular position in real value as in the upper atmosphere, and removing geometric distortions caused by changes in the scanner's speed and variations in its height, a true representation of the study site was obtained. After that the satellite image reflectance values at the sampling stations were calculated, as shown in Fig.2 and Table 2.

4-Application of Water Index (Bands Ratio Equations (
Data from the research area's multispectral bands were analyzed, and a one-way equation between the bands was used from a mathematical transaction between a defined number of bands to generate new digital images depicting specific study area attributes. In this way, water indices can be calculated and studied. In this study, 5 water indices, as shown in Table 3, were extracted using 5 equations and for 5 bands which are (band3, band4, band5, band6 and band7).   index, (ndwi)) [5] normalized difference modification the water index (mndwi) [6] moisture differential index normalized (ndmi) [7] index of automated water extraction (awei) [8] index of water ratio (wri) [9]

5-Analysis of Correlation and Regression (Statistical Analysis)
The NDWI, MNDWI, NDMI, WRI, and AWEI values of the image reflectance water index were compared to 10 distinct factors to assess the association's strength and quality. The correlation coefficient (R2) was employed as a statistical measure to assess how well the regression model described the variance of the observed data. (R2) values range from 0 to 1, with values near 1 indicating an effective regression model [10]. The satellite image bands' values and the results of laboratory analyses were correlated using the statistical tool (IBM SPSS Statistics 26); Microsoft Excel software was used to work out the study's linear mathematical equation.

Results and Discussion
The regression (R 2 ) method was used to determine the chemical and physical factors that were used, as well as to research and explain the water quality in the Al-Abbasia River. The field measurements, the date of obtaining the satellite image, and the ten sampling stations were entered into the regression equations. It appears from the values of (R 2 ) that with the exception of (PH, TDS, TSS, Na, Mg, and K), a high accuracy and a high correlation coefficient was obtained for (EC, SO 4 ,Cl,and NO 3 ) where the values of (R 2 ) were (0.691, 0.843, 0.643, and 0.612), respectively. The Pearson correlation results between water indices and parameters of water, and regression model (R 2 ) are shown in Table 4 and Table 5.

Conclusions
Some physical and chemical properties of water can be estimated using remote sensing technologies and water indices from Landsat-8 satellite photos. From the results of this study and after extracting the Regression model (R 2 ) between indices and parameters of water for the studied samples, where a high correlation coefficient was obtained between some parameters ( EC, SO 4 , Cl, and NO 3 ) and water indices (MNDWI, NDWI NDMI, AWEI, and WRI), and from results of the spectra of satellite image, it can be concluded that this correlation can be relied on in the study of these parameters in the future without the need for direct laboratory analysis. It is worth noting that the reflection in the surface of the water depends on the reflection of the sky and also on the colour of the water content depending on decomposing chemicals or living organisms at the bottom of the river, and this is the reason for making corrections and processing operations on the aerial image to obtain the real reflections without any effects.