Document Type : Research Article
Authors
1
MSc Graduated , Department of Soil Science, College of Agriculture, Ferdowsi University of Mashhad, Mashhad, Iran
2
Professors, Department of Soil Science, College of Agriculture, Ferdowsi University of Mashhad, Mashhad, Iran
3
Assistant Professor, Department of Desert and Arid Zones Management, Faculty of Natural Resources and Environment, Ferdowsi University of Mashhad, Mashhad, Iran
4
Associate Professor, Department of Desert and Arid Zones Management, Faculty of Natural Resources and Environment, Ferdowsi University of Mashhad, Mashhad, Iran
Abstract
Accurate assessment and delineation of soil salinity maps are essential for effective management and control of salinized lands. This study aimed to evaluate and map soil salinity in the Al-Suweera fields of Iraq using remote sensing and geostatistical techniques. Soil samples were collected from the topsoil layer across a 10,000-hectare area using a regular grid sampling method with 1,000-meter intervals. Electrical conductivity (EC) of the saturated paste extract was measured. Spectral bands and indices from ASTER and Landsat satellite imagery were processed in three spectral formats: digital number (DN), spectral radiance, and reflectance. Multiple linear regression was employed to establish relationships between the spectral indices and measured salinity values. Comparative analysis of the three spectral modes revealed that Landsat reflectance data provided the highest accuracy for salinity modeling (MBE = –1.24, RMSE = 4.58) among the remote sensing approaches. Furthermore, comparison between remote sensing and geostatistical methods showed that the geostatistical approach yielded superior accuracy (MBE = –0.06, RMSE = 3.53), attributed to its reliance on direct field measurements. Nonetheless, the remote sensing-derived salinity maps demonstrated acceptable accuracy and showed strong spatial agreement with the geostatistical maps.
Introduction
Soil salinity is a major environmental concern in arid and semi-arid regions, as it can negatively impact agricultural productivity and cause irreversible damage to soil. Accurate assessment of soil salinity is crucial for sustainable management and prevention of soil degradation. The Tigris alluvial plain in Iraq, an important region for crop productions, is particularly susceptible to soil salinity due to its arid climate and intensive irrigation for agriculture. Providing soil properties especially soil salinity can be used for managing the land and crop. Conventional mapping of soil and sampling are time-consuming, costly, and limited in spatial coverage. In recent years, remote sensing techniques provide an effective alternative for large-scale soil salinity mapping. The use of various spectral indices derived from satellite imagery allows for the detection of subtle changes in soil characteristics, such as soil moisture, salinity, organic matter, and texture. Geostatistics, such as kriging, is another method which offers a powerful tool for interpolating soil salinity data collected from field measurements. In recent years, numerous studies have evaluated the use of remote sensing and geostatistical methods for soil salinity mapping in different regions worldwide. However, each region has its unique characteristics that may affect the accuracy of these methods. The alluvial plain of the Tigris River in Iraq presents a unique challenge due to its specific soil properties, vegetation cover, and climate. Therefore, it is essential to evaluate the performance of these techniques in this region and identify the best approach for accurate assessment of soil salinity is thus imperative for the management and prevention of soil degradation in this region.
Material and Methods
The study was conducted in the alluvial plain of the Tigris River, in the Al-Suweera region in Alsouyreh region with a distance of 30 kilometer from Baghdad, Iraq.
Soil samples were collected using a grid sampling strategy with 1000 m distance, covering an area of 10000 ha at a depth of 0-20 cm. The collected soil samples were analyzed in the laboratory to determine the electrical conductivity (EC), pH, and sodium absorption ratio (SAR). Satellite images from the Aster and Landsat satellites were processed in three spectral modes: digital value (DN), radiance, and reflectance and then were applied to generated salinity map. The Kriging method as a geostatistics model was applied for interpolating and mapping of soil salinity, using the measured data. The estimated soil salinity maps were then compared with the measured values of soil salinity in the study area to assess the accuracy of the remote sensing and geostatistical methods. The accuracy assessment was performed using Mean Bias Error (MBE) and Root Mean Square Error (RMSE), which are commonly used metrics to evaluate the difference between estimated and known values.
Results and Discussion
The results showed that Landsat's radiance spectral mode and Aster's reflectance spectral mode were the most accurate for producing soil salinity maps. The geostatistical method, using the kriging technique, outperformed the remote sensing methods for preparing soil salinity maps. The validation results demonstrated a negative Mean Bias Error (MBE) for all remotely sensed and geostatistics methods, the lowest value (-0.06) obtained when the kriging method were applied. Additionally, kriging with the value of 3.53 yields the lowest RMSE. The results of Root Mean Square Error (RMSE) also showed a better performance of Landsat reflectance mode than all modes of Aster images. Interpolation of soil salinity using Kriging led to the most accurate maps due to the use of measured data in generating map in small area with enough taken samples. Comparison of maps generated from remote sensing data and geostatistical methods revealed almost similar salinity distribution patterns and there was a good agreement among them, demonstrating the effectiveness of remote sensing techniques which indicate the potential of remote sensing for large-scale soil salinity mapping.
The soil salinity maps generated using remote sensing and geostatistical methods both indicated lower salinity levels in the southern part of the study area compared to the northern part, which is closer to the Tigris River. This finding aligns with the actual conditions in the region, where the southern area has undergone surface drainage since 2000, effectively reducing soil salinity. In contrast, the northern part of the study area has been subject to long-term irrigation with saline water from the Tigris River (ECw=4 dS/m) and has no drainage system, resulting in persistently high salinity levels.
Conclusion
This study highlights the potential of remote sensing and geostatistical methods for mapping soil salinity in arid regions like the Tigris alluvial plain in Iraq. The findings suggest that geostatistical methods provide more accurate soil salinity maps, but remote sensing can provide valuable information at a larger scale. The produced map by both methods can provide valuable information for the management and monitoring of soil degradation in the region. The study also highlights the importance of effective drainage systems in preventing soil salinity buildup, particularly in regions where irrigation relies on salty river water.
Acknowledgements
The authors gratefully acknowledge the financial support of Ferdowsi University of Mashhad (Project Code: 47644), which made this research possible.
Keywords
Main Subjects
)
Send comment about this article