بررسی خودهمبستگی فضایی دمای سطح زمین و ارتباط آن با کاربری اراضی شهر اهواز

جزایر حرارتی شهری اثرات مخربی بر محیط‌زیست دارد و مصرف انرژی را افزایش می‌دهد. این پژوهش با هدف بررسی خودهمبستگی فضایی دمای سطح زمین و ارتباط آن با کاربری اراضی شهر اهواز انجام شد. دمای روزانه و شبانۀ سطح زمین به‌ترتیب با به‌کارگیری روش ‌پنجره مجزا از تصویر لندست 8 و محصول  مادیس mod11a1  تعیین شد. نقشۀ‌ کاربری اراضی در پنج کلاس شامل: اراضی ساخته‌شده، پوشش گیاهی، اراضی لخت، آب و رودخانۀ کارون تهیه شد. آمارۀ موران جهانی در روز و شب به‌ترتیب برابر با 0.94 و 0.9 حاصل شد که بر این اساس، دمای سطح زمین روزانه و شبانه در شهر اهواز دارای الگوی خوشه‌ای است. براساس نتایج آمارۀ گتیس- ارد جی بیشترین مساحت لکه‌های داغ روزانه و شبانه در شرق اهواز به‌ترتیب با 5050.53 و 8055.69 هکتار به دست آمد. طبق نتایج حاصل از آزمون کروسکال والیس، تفاوت معناداری بین میانگین دمای کاربری‌ها در سطح احتمال پنج درصد وجود داشت و کاربری ساخته‌شده از نوع صنعتی و اراضی لخت بیشترین دما و پهنۀ آبی و پوشش گیاهی کمترین دما را به خود اختصاص داده بودند. بین شاخص ndvi و lst همبستگی منفی (0.47-) وجود داشت. با وجود این، میانگین شاخص ndvi برابر با 0.17 بوده که نشان‌دهندۀ پوشش گیاهی ضعیف در شهر اهواز است؛ بنابراین ضرورت دارد که اقداماتی برای بهبود کیفیت پوشش گیاهی، گسترش فضای سبز شهری و حفظ منابع آبی برای تعدیل درجۀ دمای سطح زمین در شهر اهواز صورت پذیرد.

spatial autocorrelation of land surface temperature and its relationship with land use in ahvaz city

abstracturban heat islands could have destructive effects on the environment and increase energy consumption. the present study aimed to investigate the spatial autocorrelation of land surface temperature and its relationship with land use in ahvaz city. daily and night land surface temperature maps were extracted from the split window method for the landsat 8 satellite image and mod11a1 product, respectively. land use maps of the study were extracted into five classes: built-up area, vegetation, bare land, water, and karun river. based on the results, global moran’s index on day and night was equal to 0.94 and 0.9, respectively, according to which land surface temperature is distributed in a cluster form. results of getis-ord gi statistics showed that the largest area of daily and nightly hot spots in the east of ahvaz was obtained with 5050.53 and 8055.69 ha, respectively. the results of the kruskal wallis test showed that there was a statistically significant difference between the average temperature of land uses (p-value<0.05), so industrial areas and bare land had the highest temperature and the water bodies and vegetation had the lowest temperature. the results also showed that lst has a negative relationship with ndvi (r = -0.47); however, the average ndvi index in the study area was equal to 0.17 which indicated the poor quality of vegetation in ahvaz. therefore, it is necessary to take measures to improve the quality of vegetation and expand the urban green space and preserve water resources to adjust the land surface temperature in ahvaz.keywords: landsat 8, modis, split window, global moran, getis-ord gi. introductionland surface temperature (lst) is one of the essential components of climate in any region. it controls physical, chemical, and environmental processes on the earth, which are influenced by factors such as surface reflection and roughness of different types of land use/land cover, topography, geographical location, biophysical characteristics, and climate parameter changes over time. population growth and urbanization, industrial development, and human activities such as unprincipled changes in land use/land cover have created urban heat islands, which have devastating effects on the environment and the health of the community, as well as energy consumption, especially in cities with hot and dry climates. using traditional methods to estimate land surface temperature is impossible and time-consuming for all ground conditions. on the other hand, meteorological stations only measure temperature information for specific points. in contrast, due to land cover and other temperature conditions, temperatures in different regions are different from the recorded temperatures for a particular point, and this data cannot be generalized to the whole region. currently, remote sensing technology with continuity and data collection in broad ranges can be very efficient in estimating the land surface temperature. this study aimed to investigate the spatial autocorrelation of surface temperature and to identify the thermal islands of ahvaz using remote sensing and gis. materials and methodsdaily land surface temperature maps were extracted from the landsat 8 satellite image in june 2020 using a split window method. the method of comparison with the temperature of the synoptic station was used to evaluate the accuracy of the lst map. to investigate the changes in land surface temperature at night, the land surface temperature product of the modis sensor (mod11a1) was used. in the next step, the spatial autocorrelation of land surface temperature was investigated using the global moran index, and hot and cold spots were identified by getis-ord gi statistics. in the next step, land use maps of the study were extracted into five classes (built-up area, vegetation, bare land, water, and karun river) using the support vector machine classification method. to evaluate the efficiency of the classification method, the overall accuracy and kappa coefficient were used. the relationship between land surface temperature (lst) and land use type, and normalized vegetation difference index (ndvi) was investigated using the kruskal-wallis test and regression analysis, respectively. research findingsthe results showed that the difference between the air temperature recorded by ahwaz synoptic station and the air temperature obtained from landsat images is equal to 0.5 ° c, which indicates the high accuracy of the lst map. according to the results, region 8 had the highest, and region 6 had the lowest average surface temperature, equal to 53.72 and 51.53 ° c, respectively. the average lst was 51.96 ° c. the results of calculating the global moran index showed that the lst in ahvaz has a cluster pattern. the global moran index for day and night was equal to 0.94 and 0.9, respectively, which was significant at the level of one percent probability. the results of calculating the getis-ord gi statistic for identifying the areas with high and low clusters showed that daily cold spots in the west and east of ahvaz were equal to 4541.22 and 2644.74 ha, respectively. the largest area of daily and night cold spots related to region 3 is equal to 1404.27 and 344.68 ha, respectively, and the lowest area of daily cold spots related to region 7 is equal to 728.83 ha. in general, parks, areas around the karun river, cultivated lands, tree and forest cover, and water areas formed the cold spots of ahvaz city. the area of daily hot spots in the east and west of ahvaz were obtained at 5050.53 and 2022.48 ha, respectively, and the area of night hot spots in the east and west of ahvaz was obtained at 8055.69 and 4509.66 ha, respectively. the highest area of hot spots day and night is related to region eight equal to 2734.92 and 3160.65 ha, respectively. the hot spots mainly covered industrial areas and surrounding urban areas, hospitals, passenger terminals, gas stations, locations near busy roads, and bare land. based on the result, the kappa coefficient and overall accuracy of classification were 95.1% and 91%, respectively. according to the results of the kruskal-wallis test, there was a statistically significant difference between the mean temperature of land (p-value <0.05), so industrial areas and bare land had the highest average mean, i.e., the highest temperature and the water bodies and vegetation had the lowest mean rank, i.e., the lowest temperature. the result also showed that lst has a negative relationship with ndvi (r = -0.47), and with increasing the positive values of the ndvi index, the average surface temperature decreased.