background and aim: sand dunes are one of the natural effects in the western region of yazd and ashkzar region, which threaten settlements and farms, so the purpose of this research is to investigate the changes and identify sand dunes using spectral indicators in this region. in this research, 6 spectral indices ci, gsi, ndsi, bsi, ndsdi and ndsli were used to evaluate the expansion of sand dunes and their changes in the period from 2013 to 2023. from the combination of these indices and taking average of them, the asi index was introduced.
method: in this research, the extent and intensity of the expansion of sand dunes in ashkzar region using 6 spectral indices, crust index (ci), grain size index (gsi), barren soil index (bsi), normalized difference of sand dunes index (ndsdi), normalized difference sand index (ndsi) and normalized difference soil index (ndsli) were evaluated. then, the aggregate sandification index (asi) was estimated by aggregating and averaging these indices. sandy landscapes were classified into four groups of active, semi-active, semi-fixed and fixed dunes using the jenks method.
results:the results show that according to the asi index, about 73.8 square kilometers of ashkazar region were occupied by sand dunes, which is 48.2% of the entire region. from 2013 to 2023, the extent of these dunes has decreased by 4.8% and the extent of semi-active sand dunes has increased by 2.4%. the studied area was classified into 4 groups of active, semi-active, semi-fixed and fixed sand dunes using 6 spectral indices. similarities were observed among spectral indices in terms of extent and speed of movement. on the other hand, spatial and temporal patterns are very evident among the layers of sandy landscapes. the implication of these patterns is that semi-arid regions are very sensitive to climatic and human factors over a long period of time. active and semi-active sandy areas have continuously changed during the studied period and stabilized in some areas. this process has made water areas and rainfed agricultural lands in danger. based on the correlation index, it was observed that the asi index had the highest correlation with the ndsi, gsi and ci indices, in which the ndsi index and gsi had the highest correlation with values of 0.98 and 0.96. finally, kappa coefficient was used to verify the validity and capability of each index. but the kappa coefficient was calculated for the year 2023. based on this, it can be seen that the asi index has the highest kappa coefficient of 0.95 and the ndsi index is in second place with a kappa coefficient of 0.93. the lowest kappa coefficient of 0.62 belonged to the ndsdi index. after that, the ndsli index is in second place with a kappa coefficient of 0.67. kappa coefficient values were estimated based on confusion matrix in arc gis software. the findings of this research show the need for coordinated efforts to control sand desertification in the ashkazar region of yazd. the rational exploitation of the sand resources of this region and the development of sand-related industries is an important way to reverse the development process of the desert, and with the development of economic infrastructure, it increases the income of farmers and ranchers. the development of quicksand can be controlled to some extent through measures such as artificial afforestation, selection of drought-resistant plants, quicksand stabilization, fencing and soil protection.
conclusion:the results show that according to the asi index, about 73.8 square kilometers of ashkzar region were occupied by sand dunes, which is 48.2% of the entire region. from 2013 to 2023, the extent of these dunes has decreased by 4.8% and the extent of semi-active sand dunes has increased by 2.4%. based on the correlation index, it was observed that the asi index had the highest correlation with the ndsi, gsi and ci indices, in which the ndsi index and gsi had the highest correlation with values of 0.98 and 0.96. finally, based on the kappa coefficient, it was observed that the asi and ndsi index with values of 0.95 and 0.93 had the highest degree of accuracy. the lowest kappa coefficient of 0.62 belonged to the ndsdi index. after that, the ndsli index is in second place with a kappa coefficient of 0.67. kappa coefficient values were estimated based on confusion matrix in arc gis software the findings of this research show the need for coordinated efforts to control sand desertification in the ashkazar region of yazd. the rational exploitation of the sand resources of this region and the development of industries related to sand is an important way to reverse the development of the desert, and with the development of economic infrastructure, it increases the income of farmers and ranchers. through measures such as artificial afforestation, selection of drought-resistant plants, quicksand stabilization, fencing and soil protection, the development of quicksand can be controlled to some extent. also, the results showed that vegetation, pastures, irrigated lands, rainfed agricultural lands, settlements and infrastructures are at high risk of desertification.