| چکیده انگلیسی مقاله |
Alteration of urban forms and geometries due to rapid unplanned urban development can result in localized elevation of land surface temperature - a phenomenon known as urban heat island. This study examins the impact of land development patterns on land surface temperatures in a heterogeneous urban environment. As many as 18 Landsat satellite images for 1995, 2008 and 2021 (average cloudless images per year) were used in this sutdy. First, land development patterns were generated from the processing of Landsat satellite images in SAGA GIS using the method of Stewart and Oke. At the second stage, land surface temperatures for 1995, 2008, and 2021 were extracted using the single-channel algorithm. At the third stage, the relationship between mean value of land surface temperatures and the heterogeneous form of Tehran was analyzed. The results showed that among all LCZs, the highest mean value of land surface temperatures belonged to the Heavy Industry LCZ with the average temperatures of 5.32, 48.18, and 51.87°C for 1995, 2008, and 2021, followed by the Bare Soil/Sand LCZ with the average temperatures of 47.25, 49.25 and 52.36°C for the same years. The lowest mean value of land surface temperatures belonged to the Water LCZ with the average temperatures of 20.5, 22.63, and 23.15°C for 1995, 2008, and 2021, respectively. It was also found that the Compact Low-Rise LCZ had higher mean value of land surface temperatures than both Compact High-Rise and Compact Mid-Rise LCZs. The differences observed between the highest and lowest land surface temperatures in the study area are indicative of the significant impacts of urban form on land surface temperatures. The findings can contribute to our understanding of urban environments and help to devise better plans for minimizing the adverse effects of land use and development on these environments. Extended Abstract 1-Introduction Rapid unplanned urban development tends to result in reduced surface albedo, decreased vegetation, increased anthropogenic heat generation due to production and transportation, and altered urban geometry, which can all play a major role in the emergence of Urban Heat Islands (UHIs). Over the years, a variety of models and metrics have been developed for the assessment of UHIs, but there is still much research to be done in this area. In the great majority of previous studies in this field, UHIs have been assessed on a macro scale using urban-rural, urban-agricultural, urban-suburban, urban-nonurban and urban-water temperature differences. However, since temperature distribution over an urban area depends on a wide range of factors, modeling outputs such as Land Surface Temperature (LST) estimates can become unrealistic if these factors are not taken into account. Thus, such assessments need to be made on a local scale with more attention to details. Among the variety of models proposed for UHI assessment, the one called Local Climate Zone (LCZ) has managed to address many issues of its predecessors. LCZ stands out among models of this kind for its excellent classification, capability, and precision, and provides a relatively new approach to the climatic classification of urban and non-urban areas. Much of the research done on UHIs has been focused on thermal conditions rather than urban form. This is while urban form, open spaces, and building dimensions can all affect the rate of energy absorption/reflection, and therefore, fossil energy consumption and LST. However, so far, not much research has been conducted on the impact of urban morphology, building dimensions and other urban form factors on the intensity of UHI, especially in Iran. To fill this gap, this study investigated the effect of urban form on LST and UHI in order to gain more insight into how it impacts urban temperature. 2-Materials and Methods In this study, first, the study area and the factors related to urban form and land cover were examined, and LCZ classification was performed in Saga GIS based on the model’s 17 standard classes. Then, LST was estimated using the single channel algorithm based on satellite spatio-temporal data and the level of LST and UHI in different LCZ classes was examined. Finally, the relationship between LCZ classes and spatio-temporal changes of UHI and NDVI in 1995, 2008 and 2021 was investigated. Given the impact of climatic and seasonal conditions on LST, only cloud-free images were used in LST modeling. In the end, the mean value of LST for each year was obtained by averaging all LSTs obtained from all cloud-free images taken over that year. 3- Results and Discussion The LST estimates obtained for hot and cold seasons of 1995, 2008 and 2021 showed that the northern parts of the study area have a low LST, which can be attributed to their loosely packed urban structure as well as green spaces. Natural covers like dense orchards and water bodies can also have very low LSTs because of evapotranspiration. The central parts of the city were found to have high LSTs because of high building density and poor green cover. The results also showed a steady decline in the area of “cold” and “very cold” LST classes over the past 16 years. An inverse correlation was observed between NDVI and normalized LST, in the sense that LST was higher wherever vegetation was lower. The lowest LSTs belonged to areas with the highest levels of vegetation. Also, LST was found to be dropping with movement toward more vegetated areas. The 1995, 2008 and 2021 mean value of LSTs obtained for LCZs showed a steady rise in the average temperature of all zones. Over these years, the average temperature in “heavy industry”, “large low-rise”, “sparsely built”, “open low-rise”, “compact low-rise”, “compact mid-rise” zones has risen from 30-45°C in 1995 to 31-47°C in 2008 and to 33-5°C in 2021, which has a notable ascending trend. Of all LCZs, “heavy industry” had the highest LST and “water” the lowest LST for all three years. Also, all LCZs with vegetation land use (green cover) have exhibited rising temperatures over the study period. 4- Conclusion In this study, LCZ classification was performed using satellite imagery and telemetry data. A total of 17 classes, including 10 built classes and 7 natural cover classes were identified in the study area. LST in three periods from 1995 to 2021 was extracted from the Operational Land Imager (OLI) data using the single-channel algorithm. The results demonstrated a relationship between urban form and LST and UHI. The highest overall accuracy and kappa coefficient for LCZ classification were for 1995 with Acc=93.58% and κ=0.91 and the lowest were for 2018 with Acc=89.56% and κ=0.87. |