Document Type : Research Paper
Authors
1
PhD in Desert Management and Control, Combat to Desertification Department, Faculty of Desert Studies, Semnan University, Semnan, Iran
2
Corresponding Author, Professor, Combat to Desertification Department, Faculty of Desert Studies, Semnan University, Semnan, Iran
3
Assistant Professor, Iranian Space Research Institute, Tehran, Iran
4
M.Sc. in Remote Sensing, Iranian Space Research Institute, Tehran, Iran
5
Assistant Professor, Agricultural Education and Extension Institute, Agricultural Research, Education and Extension Organization (AREEO), Tehran, Iran
10.22092/ijfrpr.2025.368227.1657
Abstract
Background and Objectives: Global warming can influence soil health and productivity by accelerating the rate and intensity of erosion. Previous studies have shown that arid and semi-arid regions are at higher risk of wind erosion than other regions due to increased soil surface evaporation and reduced soil moisture. Temperature variations directly affect soil moisture and crop production in plants. For example, high temperatures cause soil drying, which limits plant growth. Drier soils with less vegetation cover are more prone to wind erosion. Furthermore, temperature changes modify the pressure gradient in desert and arid regions, leading to strong and persistent winds, which are among the main causes of dust storms in these areas.
Methodology: This study examined the relationship between the frequency of dust storms and land surface temperature (LST), mean temperature, maximum temperature (Tmax), minimum temperature (Tmin), and vegetation cover in Sistan-Baluchistan Province. These variables were derived from MODIS satellite imagery and meteorological data over a 20-year period (2000–2020). The DSI index was used to analyze the trend of dust storm days. A multilayer Perceptron neural network (MLP) was employed to predict the importance of the studied variables in the frequency of dust storm days. In designing the neural network, the input activation function was sigmoid, while the output activation function was identity. The network was structured with one hidden layer and five neurons. Linear regression was also used to evaluate the trends of variable changes.
Results: The results showed that the trend of changes in the DSI index in the study area was decreasing and statistically significant (P-value = 0.007). The highest and lowest values of this index were 40 and 4, respectively. The LST index and Tmax were the most important variables, while NDVI had the least importance in predicting dust storms in the study area. The highest and lowest values of the LST index were 10°C and 6°C, respectively. The trend of this index has decreased significantly in recent years (P-value = 0.03). Spatially, the highest LST values were recorded at Zabol, Iranshahr, and Saravan stations, while the lowest values were observed at Zahedan, Khash, and Chabahar. The highest and lowest Tmax values were 32°C and 30°C, respectively, and its trend was not statistically significant.
Conclusion: Among the temperature-related variables studied, LST and Tmax had the greatest impact on changes in the dust storm index. In the study area, the average LST and Tmax were measured at 8°C and 31°C, respectively, which can strongly influence other environmental factors such as air pressure, wind, and soil moisture. In contrast, NDVI, mean temperature, and Tmin had the least effect. The spatial variation pattern also confirmed the influence of LST and Tmax on DSI, with the northern part of Sistan-Baluchistan Province experiencing a more critical situation in terms of dust storm frequency. In conclusion, considering the key role of land surface temperature in DSI changes, management and planning in Sistan-Baluchistan should focus on appropriate land use planning and controlling land use changes to reduce surface temperature and limit dust storm sources.
Keywords
Main Subjects
)