Effect of land cover changes on land surface temperature in Yazd plain, Iran

The growth of urbanization and the expansion of cities have caused significant changes in the surrounding land cover. Changes in land surface characteristics as a result of rapid urbanization can affect climatic conditions in seasonal and long-term periods on a local, regional and global scales. Land surface change (LST) change is directly related to land cover changes, so that land cover change causes changes in the land surface temperature. Yazd plain as the study area, has witnessed many changes in land cover during the past two decades due to immigrants, so studying its temperature changes is very important. The aim of this study was to investigate the changes in land surface temperature of the Yazd plain over an 18-year period and compare it with land cover changes in order to determine the effect of these land cover changes on land surface temperature. The distinction between the current study and previous studies is the effect of land cover change on land surface temperature in different months of the year.

In this study, Landsat 5 and 8 satellite images were used to investigate the effect of land cover changes on land surface temperature in 2001 and 2019. Land surface temperature map were estimated for Landsat 5 satellite images using single band method, and for Landsat 8 images using split-window algorithm. The land cover maps for the studied years were prepared using the supervised classification method of maximum likelihood. Changes in land cover made from 2001 to 2019 and its effect on land surface temperature were examined.

In the land cover map of the study area, six land cover types of mountainous rangelands, plain rangelands, agriculture, sand dunes, barren lands and residential areas were identified. For better separation of land cover, images related to the peak of vegetation period were used in the classification processes. After determining the degree of class segregation, the supervised classification was performed by the maximum likelihood method, and then land cover changes were detected in the area. In order to investigate the changes made in the study area, the area and changes of each land cover were also extracted. Urban areas with an increase of 12.1% area had the highest growth, and agricultural lands with 28% reduction had the highest reduction in area. At the same time, the study period with increasing migration from villages to urban areas, the phenomenon of urbanization and industrialization of the study area can be the reason for this issue. Also, the 24.5% increase in the area of sand dunes is due to recent droughts and the placing of the study area in the sedimentation site of the wind erosion process. Although little vegetation in sand dunes has stabilized them, in recent years with the loss of this cover, the area of wind erosion process has increased. Land cover changes map was extracted from the area in 18-year period. The Land surface temperature maps for four months of the year were prepared using the single channel method in 2001, and the split window algorithm in 2019. For better comparison, the relevant Land surface temperature indexes per month was classified into 10 categories at 2 °C intervals and the areas of the classes were compared with each other. In order to investigate the effect of land cover change on land surface temperature, the mean, minimum and maximum land surface temperatures in each land cover on the mentioned dates were examined. The results show that the land surface temperature in each area of the earth is influenced by surface factors and its characteristics and the temperature obtained in different land cover is different. Land surface temperatures in bare lands and sand dunes that have been converted into other land covers have dropped significantly, so that, in July and September it dropped about 2°C. Most of the temperature increase is related to the conversion of agricultural lands to bare lands, sand dunes and mountainous rangelands. Contrary to the existing hypotheses that the greatest increase in temperature is related to changing agricultural to residential land cover, in this area located in arid regions of central Iran, the conversion of agricultural lands to bare lands and sand dunes has led to higher temperatures. The change in land surface temperatures in the changed coverage in February was higher than in other months and less in September.

The results of this study showed that during 2001-2019, the area of residential lands and sand dunes increased and the area of agricultural lands, bare lands, and rangelands decreased. The results of land surface temperature assessment in changed land cover showed that the land surface temperature in bare lands and sand dunes that have been converted to other coatings has decreased significantly. Conversion of agricultural lands to bare lands and sand dunes has led to higher temperatures than other land cover type and physical development has created numerous environmental problems.