|
|
|
|
 |
Volume 18, Issue 66 (8-2024) |
 |
|
|
|
|
Investigating the Impact of Climatic and Land Surface Characteristics on Groundwater Quality Using Google Earth Engine (GEE) Data (Case Study: Kahorestan Plain, Hormozgan)
|
Adnan Sdeghi lari *   , Mohamad Kazemi , Fatemeh Rajabi |
|
|
|
Abstract: (1095 Views) |
Introduction
Groundwater, as one of the vital resources for supplying drinking water and agriculture, is significantly affected by climate change and human activities. Rising temperatures, changing precipitation patterns, and alterations in vegetation lead to changes in the chemical composition of water, which can have serious implications for the provision of drinking water and agricultural needs. The quality of groundwater, like that of surface water, is constantly changing, and in arid and semi-arid plains, identifying and analyzing the factors influencing groundwater quality is crucial for managing various quality zones. Researchers in this field have acknowledged that soil characteristics and land surface can directly impact groundwater quality, and proper soil management can contribute to improving water quality. Therefore, this study examines the impact of climatic variables and land surface characteristics on groundwater quality in the Kahorestan plain of Hormozgan Province, aiming to identify quality changes in water during irrigation and non-irrigation seasons. It is noteworthy that previous research primarily estimated the Water Quality Index (WQI) using interpolation algorithms within Geographic Information Systems (GIS) and delineated different groundwater quality levels. In the present study, in addition to investigating the significant relationships between climatic variables and land surface characteristics with groundwater quality, the spatial distribution of these variables within the study area is analyzed, and modeling is conducted using Geographically Weighted Regression (GWR) and Ordinary Least Squares (OLS) methods based on the influential independent variables.
Methodology
The study area in the present research is the Kahorestan plain, covering an area of 273.09 square kilometers, which is one of the rare plains in the western part of Hormozgan Province. This study was conducted based on chemical data from 15 semi-deep wells in the Kahorestan plain during irrigation and non-irrigation periods in the years 2009 and 2018. The independent variables of the study included the Normalized Difference Salinity Index (NDSI), cumulative precipitation (Pr), Normalized Difference Vegetation Index (NDVI), soil water deficit (DEF), potential evapotranspiration (PET), land surface temperature (LST), Palmer Drought Severity Index (PDSI), and actual evapotranspiration (AET), which were extracted as monthly raster satellite images from the Google Earth Engine (GEE). The dependent variable of the study was the Groundwater Quality Index (WQI), which was modeled within a Geographic Information System (GIS). To assess the significance of the relationships between these variables at different time periods, the Chi-square test was employed. Subsequently, Geographically Weighted Regression (GWR) and Ordinary Least Squares (OLS) techniques were utilized to model the spatial variations in groundwater quality. To compare the validity or efficiency of the multiple regression models, criteria such as the coefficient of determination, root mean square error, and Akaike Information Criterion were used. Additionally, it is noted that the Local Moran's I statistic was employed to investigate the spatial distribution of groundwater quality zones during these time periods.
Results and discussion
The chi-square test indicated a significant relationship between climatic variables and land surface characteristics with groundwater quality. Spatial autocorrelation analysis revealed that the spatial distribution of groundwater quality in the Khorasan plain is random. Among the two methods, GWR (Geographically Weighted Regression) and OLS (Ordinary Least Squares), the GWR technique provided better results with a root mean square error of 0.14, a residual sum of squares of 11.3, a coefficient of determination (R²) of 0.82, and an Akaike Information Criterion (AIC) of -570.19, compared to the OLS method. Additionally, NDVI (Normalized Difference Vegetation Index) and NDSI (Normalized Difference Snow Index) showed a coefficient of determination (R²) of 0.47 in the first time period (4/2009), with the NDSI variable showing an R² of 0.63 in the second time period (11/2009). In the third time period (5/2018), both NDVI and NDSI had an R² of 0.65, and in the fourth time period (12/2018), the two variables NDVI and NDSI again showed the highest impact among climatic variables and land surface characteristics on groundwater quality, with an R² of 0.55. The groundwater quality classes of the Kahorestan plain only included two categories: very poor and unsuitable for agricultural use. In the first-time frame, the "very poor" quality class of groundwater in the Khorasan plain covered an area of 2.44 Km2. In the second time frame, the only groundwater quality class in the Khorasan plain during that irrigation season was "unsuitable for agricultural use," with an area of 275.2 Km2. In the third time frame, May 2018, the "very poor" quality class of surface groundwater accounted for 14.91 Km2 of the plain's area. In the fourth time frame, December 2018, the area classified as "very poor" quality was estimated to be 8.51 Km2.
Conclusion
As the results indicated, groundwater quality during irrigation seasons (months 11 and 12 of the years 2009 and 2018) decreases compared to non-irrigation seasons (months 4 and 5 of the years 2009 and 2018), with the quality levels of groundwater also diminishing during irrigation periods. The findings revealed that increased vegetation cover and reduced soil salinity are key factors in improving groundwater quality in this region. A decrease in vegetation cover and an increase in soil salinity during irrigation seasons lead to a reduction in the extent and quality of good and moderate quality zones. Given the superiority of the GWR model over OLS, this model can serve as a useful tool in managing the quality zones of groundwater resources and analyzing changes in water quality. Ultimately, this research emphasizes that attention to climatic variables and land surface characteristics is essential for the optimal management of groundwater resources. This study not only utilized various interpolation methods for the WQI based on the chemical data from observation wells within a Geographic Information System (GIS), but also highlighted the prediction of different groundwater quality zones using Geographically Weighted Regression and Ordinary Least Squares techniques, considering the impact of various climatic variables and land surface characteristics. Additionally, obtaining independent variables at a monthly scale across different time periods of irrigation and non-irrigation seasons from the online Google Earth Engine (GEE) was another focus of the current study regarding the examination of groundwater quality changes. |
Article number: 6 |
| Keywords: Soil salinity index, groundwater quality, climatic variables, geographically weighted model, Global Moran's I. |
|
|
Full-Text [PDF 1778 kb]
(560 Downloads)
|
Type of Study: Research |
Subject:
Special
Received: 2024/08/6 | Accepted: 2024/08/31 | Published: 2024/09/15
|
|
|
|
|
|
|
| Send email to the article author |
|
|
|
| Add your comments about this article |
|
|
|
|
Sdeghi lari A, Kazemi M, Rajabi F. Investigating the Impact of Climatic and Land Surface Characteristics on Groundwater Quality Using Google Earth Engine (GEE) Data (Case Study: Kahorestan Plain, Hormozgan). jwmseir 2024; 18 (66) : 6 URL: http://jwmsei.ir/article-1-1171-en.html
|
|
|
|
|
|
|
