Determining the desertification intensity based on spectral indices using Sentinel-2 images (Case study: Sistan and Baluchestan province)

Subject Areas : Natural resources and environmental management

Farhad Zolfaghari ¹ , Vahideh Abdollahi ²

1 - Assistant Professor, Higher Education Complex of Saravan, Iran
2 - Assistant Professor, Higher Education Complex of Saravan, Iran

Received: 2021-05-14 Accepted : 2021-08-10 Published : 2022-04-21

Keywords: Desertification, Sentinel-2, Spectral index, Albedo, Topsoil Grain Size Index (TGSI) and Normalized Difference Vegetation Index (NDVI),

Abstract :

Background and Objective Different vegetation covers have different albedo levels. On the other hand, surface albedo is one of the most important components of surface radiation balance, which can be used to identify severely degraded and desertified regions. Vegetation can be considered as one of the most important key components in arid regions to reduce the effects of erosion and desertification due to the effects of vegetation for land surface stability. Expansion of desertification and also changes in vegetation cover, could be change the surface Albedo. The purpose of this study is to determine the desertification intensity based on spectral indices, Albedo, Topsoil Grain Size Index (TGSI) and Normalized Difference Vegetation Index (NDVI) using remote sensing technology. Identification the damaged areas with the lowest cost in the shortest time, using Sentinel-2 images with a spatial resolution of 10 meters is one of the objectives of this study. Also, this study will introduce the best indicator for monitoring desertification intensity in arid regions for the first time in the Sistan and Baluchestan region based on spectral indices using Sentinel-2 images.Materials and Methods The following steps were performed to evaluate the intensity of desertification and identify the appropriate indicator in order to mapping the desertification intensity: 1) Selection the images and perform image preprocessing operations using SNAP software; 2) Calculation of TGSI, NDVI and Albedo indices; 3) Investigation the correlation between indices using SPSS®24 software. 4) Preparation of desertification intensity map of the region and obtaining the equation of desertification intensity using ArcGIS®10.3 software. In the first step of this research, Sentinel-2A satellite data related to MSIL-1C sensor was selected on August 20, 2020. The images were selected in such a way that the growing season of the plants is not annual and temporary, and also the day was selected when there is no cloud cover. The required images were downloaded and used from the URL address: http://scihub.copernicus.eu/. Results and Discussion The results of linear regression between NDVI and Albedo indices showed that, these two indices had negative correlation, and the correlation coefficient in Souran and Zabol was 0.76 and 0.63, respectively. The results showed that with increasing NDVI, decreased of the albedo index occurred. Also, the results of linear regression model showed strong and positive relationship between TGSI and Albedo indices, as the correlation coefficient of Souran and Zabol was 0.78 and 0.81, respectively. The results showed that the TGSI and the albedo simultaneously decreased or increased. Desertification intensity in the study areas was determined based on the equation I= a × Index ± Albedo and also by using Natural Breaks (Jenks) method in ArcGIS software, desertification intensity of study areas classified to 5 degrees, 1. Without desertification, 2. Low desertification, 3. Moderate desertification, 4. Severe desertification, and 5. Extremely desertification. In this study Albedo, NDVI and TGSI indices were extracted based on Sentinel-2 satellite data. The results of linear regression between NDVI and Albedo showed that there is strong negative relationship between these indices that was consistent with the results of similar studies. The high and negative correlation, means that any increase in the vegetation cover will lead to decrease the Albedo. On the other hands the areas with high Albedo, indicate degradation of vegetation cover and bare soil. In the regions with sever desertification intensity, the value of surface Albedo was high and the vegetation cover was low. Classification of desertification intensity in Sistan region based on Albedo-NDVI model showed that 27.73% of the area were in the class of without desertification intensity, 18.03% in the low class, 32.92% in the moderate class, 20.3% were in the severe class and only 1.02% of the area were in the very severe desertification intensity class. Also, the classification of desertification intensity in Souran based on Albedo-NDVI model showed 4.82% of the area without desertification, 8.44% in low class, 50.97% in moderate class, 34.48% in severe class and 1.3% of the area were in very severe desertification class. The highest percentage of desertification intensity of the area were in the moderate class. The results of linear regression between TGSI and Albedo indices also showed that there is a positive and strong relationship between these indices. The results showed that the relationship between TGSI and Albedo indices was stronger than the relationship between NDVI and Albedo indices and in both regions the correlation coefficient was higher. One of the main reasons for this is the dispersion of vegetation cover in arid areas. The relationship between TGSI and Albedo better shows the spatial characteristics of vegetation-free areas as well as areas with very low vegetation cover to determine the intensity of desertification. The TGSI index reflects the coarse particle size of the topsoil, which has a positive relationship with the fine sand content of the topsoil. Whatever the larger particle size of the topsoil, will have the greater desertification intensity. In the areas where the content of fine sand in the topsoil is high, the high range of TGSI index will be seen.Conclusion In this study, using Sentinel-2 multispectral images and remote sensing technique, we extracted the intensity of desertification in different arid regions of the Sistan and Baluchestan province, for the first time in Iran. Based on the spectral reflection that occurred from the ground and the spatial resolution of 10 meters, we studied the intensity of desertification in two areas. Based on the results of this research, we suggest to use the combination of Albedo-NDVI and Albedo-TGSI models in order to monitoring the desertification intensity in arid regions of Iran. The results of this study showed that areas without desertification and low intensity of desertification are better identified based on Albedo-TGSI model.

References:

Ait LA, Saber H, Pradhan B. 2018. Quantitative assessment of desertification in an arid oasis using remote sensing data and spectral index techniques. Remote Sensing, 10(12): 1862. doi:https://doi.org/10.3390/rs10121862.

Allen R, Tasumi M, Trezza R. 2002. Surface Energy Balance Algorithms for Land. Advanced Training and User’s Manual Idaho Implementation, 240 p.

Bernardo SBd, Braga AC, Braga CC, de Oliveira LM, Montenegro SM, Barbosa Junior B. 2016. Procedures for calculation of the albedo with OLI-Landsat 8 images: Application to the Brazilian semi-arid. Revista Brasileira de Engenharia Agrícola e Ambiental, 20: 3-8. doi:https://doi.org/10.1590/1807-1929/agriambi.v20n1p3-8

Cai G, Du M, Liu Y. 2011. Regional Drought Monitoring and Analyzing Using MODIS Data — A Case Study in Yunnan Province. In, Berlin, Heidelberg,  Computer and Computing Technologies in Agriculture IV. Springer Berlin Heidelberg, pp 243-251. doi:https://doi.org/210.1007/1978-1003-1642-18336-18332_18329.

Cordeiro MC, dos Santos NA, Silva VMA, de Melo Luiz D, da Silva VdPR. 2015. Case study: identification of desertification in the years 1999, 2006 and 2011 in Mossoró-RN. Journal of Hyperspectral Remote Sensing, 5(4): 101-106. doi:https://doi.org/10.29150/jhrs.v5.4.p101-106.

Eftekhari R, Shahriyari A, Ekhtesasi M. 2015. Assessment and mapping of current and potential of desertification using MICD Model with emphasis on wind erosion criteria in southwest of Hirmand city. Journal of Development and Geography, 38: 139- 150. (In Persian).

Eskandari S. 2019. Comparison of different algorithms for land cover mapping in sensitive habitats of Zagros using Sentinel-2 satellite image:(Case study: a part of Ilam province). RS & GIS for Natural Resources 10(1): 72-86. (In Persian).

Fozuni L. 2007. Evaluation of the current status of desertification Sistan plain using modify MEDALUS Model with emphasis on wind and water erosion criteria. Master degree of desertification, University of Zabol. 215 p. (In Persian).

Gillespie TW, Ostermann-Kelm S, Dong C, Willis KS, Okin GS, MacDonald GM. 2018. Monitoring changes of NDVI in protected areas of southern California. Ecological Indicators, 88: 485-494. doi:https://doi.org/10.1016/j.ecolind.2018.01.031.

Goudie AS, Middleton NJ. 2006. Desert dust in the global system. Springer Science & Business Media. 288 p.

Han L, Zhang Z, Zhang Q, Wan X. 2015. Desertification assessments in the Hexi corridor of northern China’s Gansu Province by remote sensing. Natural Hazards, 75(3): 2715-2731. doi:https://doi.org/10.1007/s11069-014-1457-0.

Houldcroft CJ, Grey WM, Barnsley M, Taylor CM, Los SO, North PR. 2009. New vegetation albedo parameters and global fields of soil background albedo derived from MODIS for use in a climate model. Journal of Hydrometeorology, 10(1): 183-198. doi:https://doi.org/10.1175/2008JHM1021.1.

Jahantigh M, Jahantigh M. 2020. Study effect of flood productivity on vegetation changes using field work and Landsat satellite images (Case study: Shandak of Sistan region). RS & GIS for Natural Resources 10(4): 57-73. (In Persian).

Kaffash A, Rouhimoghadam E, Afshari A, Zolfaghari F. 2018. Investigation the effects of Climate, Vegetation, Wind Erosion and Soil Criteria on desertification Potential Using GIS (Case Study: Moradabad Saravan Regio). Journal of Geographical New Studies Architecture and Urbanism, 2(14): 15-29. (In Persian).

Kang HS, Hong SY. 2008. An assessment of the land surface parameters on the simulated regional climate circulations: The 1997 and 1998 east Asian summer monsoon cases. Journal of Geophysical Research: Atmospheres, 113(D15). doi:https://doi.org/10.1029/2007JD009499.

Kariminazar M, Mosaaedi A, Moghadamnia A. 2010. Investigation of climatic factors affecting occurrence of drought (Case Study of Zabol Region). Journal of Water and Soil Conservation, 17(1): 145- 158. (In Persian).

Karnieli A, Qin Z, Wu B, Panov N, Yan F. 2014. Spatio-temporal dynamics of land-use and land-cover in the Mu Us sandy land, China, using the change vector analysis technique. Remote Sensing, 6(10): 9316-9339. doi:https://doi.org/10.3390/rs6109316.

Lamchin M, Lee J-Y, Lee W-K, Lee EJ, Kim M, Lim C-H, Choi H-A, Kim S-R. 2016. Assessment of land cover change and desertification using remote sensing technology in a local region of Mongolia. Advances in Space Research, 57(1): 64-77. doi:https://doi.org/10.1016/j.asr.2015.10.006.

Lamchin M, Lee W-K, Jeon SW, Lee J-Y, Song C, Piao D, Lim CH, Khaulenbek A, Navaandorj I. 2017. Correlation between desertification and environmental variables using remote sensing techniques in Hogno Khaan, Mongolia. Sustainability, 9(4): 581. doi:https://doi.org/10.3390/su9040581.

Myhre G, Myhre A. 2003. Uncertainties in radiative forcing due to surface albedo changes caused by land-use changes. Journal of Climate, 16(10): 1511-1524. doi:https://doi.org/10.1175/1520-0442(2003)016<1511:UIRFDT>2.0.CO;2.

Naegeli K, Damm A, Huss M, Wulf H, Schaepman M, Hoelzle M. 2017. Cross-comparison of albedo products for glacier surfaces derived from airborne and satellite (Sentinel-2 and Landsat 8) optical data. Remote Sensing, 9(2): 110. doi:https://doi.org/10.3390/rs9020110.

Pan J, Li T. 2013. Extracting desertification from Landsat TM imagery based on spectral mixture analysis and Albedo-Vegetation feature space. Natural Hazards, 68(2): 915-927. doi:https://doi.org/10.1007/s11069-013-0665-3.

Parvariasl H, Pahlavanravi A, Moghaddamnia A. 2010. Assessing desertification hazard in Neiyatak region using ESAs Model. Journal of Iran Natural Resources, 2: 42- 54. (In Persian).

Piña RB, Díaz-Delgado C, Mastachi-Loza CA, González-Sosa E. 2016. Integration of remote sensing techniques for monitoring desertification in Mexico. Human and Ecological Risk Assessment: An International Journal, 22(6): 1323-1340. doi:https://doi.org/10.1080/10807039.2016.1169914.

Scott D, Smart M. 1999. Wetlands of the Sistan Basin, South Caspian and Fars, Islamic Republic of Iran, Ramsar Convention Monitoring Procedure Report No.26. 110 p.

Wei H, Wang J, Cheng K, Li G, Ochir A, Davaasuren D, Chonokhuu S. 2018. Desertification information extraction based on feature space combinations on the Mongolian plateau. Remote Sensing, 10(10): 1614. doi:https://doi.org/10.3390/rs10101614.

Wei H, Wang J, Han B. 2020. Desertification information extraction along the China–Mongolia railway supported by multisource feature space and geographical zoning modeling. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 13: 392-402. doi:https://doi.org/10.1109/JSTARS.2019.2962830.

Wulder MA, Hilker T, White JC, Coops NC, Masek JG, Pflugmacher D, Crevier Y. 2015. Virtual constellations for global terrestrial monitoring. Remote Sensing of Environment, 170: 62-76. doi:https://doi.org/10.1016/j.rse.2015.09.001.

Xiao J, Shen Y, Tateishi R, Bayaer W. 2006. Development of topsoil grain size index for monitoring desertification in arid land using remote sensing. International Journal of Remote Sensing, 27(12): 2411-2422. doi:https://doi.org/10.1080/01431160600554363.

Zolfaghari F, Shahriyari A, Fakhireh A, Rashki A, Noori S, Khosravi H. 2011. Assessment of desertification potential using IMDPA model in Sistan plain. Watershed Management Research (Pajouhesh & Sazandegi), 91: 97-107. (In Persian).

Zongyi M, Xie Y, Jiao J, li L, Wang X. 2011. The Construction and Application of an Aledo-NDVI Based Desertification Monitoring Model. Procedia Environmental Sciences, 10: 2029-2035. doi:https://doi.org/10.1016/j.proenv.2011.09.318.