استخراج و تحلیل زمانی و مکانی دمای سطح زمین نسبت به متغیرهای طبیعی و انسانی توسط روش آماری نسبت فراوانی (fr) (مطالعۀ موردی: محدودۀ شرقی استان قزوین)

افزایش دمای سطحی زمین و یا ایجاد جزایر حرارتی روی سطح شهرها یکی از عوامل زیست‌محیطی است که محققان در دهه‌های اخیر به آن توجه کرده‌اند. هدف از پژوهش حاضر تحلیل مکانی و زمانی دمای سطحی بخش شرقی استان قزوین در دو مرحله است. در مرحلۀ اول پس از محاسبۀ دمای سطحی با استفاده از تصاویر ماهواره‌ای لندست 8 در سال‌های 2016، 2019 و 2021 روند تغییرات دما در تابستان و زمستان مطالعه شد. هدف از مرحلۀ دوم ارزیابی و تحلیل تاثیر برخی عوامل مختلف توپوگرافی و انسانی بر تغییرات دمای منطقه است. این ارتباط با روش آماری – مکانی نسبت فراوانی استخراج و تحلیل شد. نتایج نشان داد که در فصل تابستان وسعت نواحی با دمای بیش از 35 درجه از حدود 6600 کیلومتر مربع در سال 2016 به بیش از 9300 کیلومتر مربع در سال 2021 رسیده است. در‌مقابل، در فصل زمستان مساحت دمای زیر صفر درجه در سال 2016 از حدود 1330 کیلومتر مربع به حدود 890 کیلومتر مربع در سال 2021 رسیده است؛ بنابراین به‌طور کلی، دمای منطقه در‌طول دورۀ مد‌نظر افزایش داشته است. همچنین، از‌میان عوامل طبیعی انتخاب‌شده، ارقام محاسبه‌شدۀ نسبت فراوانی لایۀ جهت ‌شیب نشان‌دهندۀ تاثیرگذاری بیشتر این عامل نسبت به دیگر عوامل طبیعی بوده است. از طرف دیگر، پوشش اراضی بایر با رقم نسبت فراوانی 0.75 در کلاس دمایی بیش از 35 درجه سانتیگراد در فصل تابستان و پوشش برفی با رقم نسبت فراوانی 0.89 در کلاس دمایی زیر منفی 13 درجه در فصل زمستان نشان‌دهندۀ تاثیر نوع پوشش و کاربری زمین در تغییرات دمای محیط است. با استخراج چنین روابط مکانی – زمانی از منطقۀ مطالعه‌شده می‌توان اقدام‌های تاثیرگذاری را در‌زمینۀ مدیریت شهری، زیست‌محیطی و بحران انجام داد و سپس از پیامدهای منفی زیست‌محیطی بسیاری جلوگیری کرد.

spatiotemporal assessment of land surface temperature (lst) based on natural and human parameters using frequency ratio (fr) technique(case study: east qazvin province)

abstractin recent decades, one of the environmental factors that has garnered significant attention from researchers is the increase in land surface temperature (lst) and the emergence of urban heat islands in cities. this study aimed to analyze the spatial and temporal variations in lst in the eastern part of qazvin province across two phases. in the first phase, lst was calculated using landsat 8 satellite images from 2016, 2019, and 2021, allowing for an examination of temperature trends during summer and winter. the second phase focused on evaluating the influence of various topographical and anthropogenic factors on temperature changes within the region. the relationships were extracted and analyzed using the statistical-spatial frequency ratio (fr) technique. the results indicated that during summer, areas experiencing temperatures above 35°c had increased from approximately 6,600 km² in 2016 to over 9,300 km² in 2021. conversely, in winter, the extent of sub-zero temperatures had decreased from about 1,330 km² in 2016 to around 890 km² in 2021. overall, this suggested a general increase in regional temperatures over the study period. among the selected natural factors, the fr values for the aspect layer indicated it was more influential than other factors. additionally, barren land cover exhibited an fr value of 0.75 in areas with temperatures exceeding 35°c in summer, while snow cover showed a frequency ratio of 0.89 in regions with temperatures below -13°c in winter. these findings underscored the impacts of land use and cover on lst. by extracting such spatial-temporal relationships from the studied area, we could take effective measures in urban planning, environmental management, and crisis response, thereby mitigating numerous negative environmental consequences. keywords: land surface temperature (lst), spatiotemporal analysis, frequency ratio (fr), geographic information system (gis), qazvin province. introductionin recent decades, the increase in land surface temperature (lst) has become a significant environmental concern for researchers in urban and environmental studies. numerous studies have demonstrated that urban areas typically experience higher temperatures than their surrounding environments due to human activities. these activities include heating and cooling processes, transportation and road construction, industrial operations, and excessive absorption of solar radiation by urban structures. additionally, the scarcity of green spaces and poor air circulation exacerbate this issue. today, lst calculations are primarily conducted using satellite imagery, particularly through the processing of thermal infrared bands. one of the most commonly utilized datasets in this context is the thermal bands from landsat 8 satellite images (tirs). researchers have developed various algorithms for calculating lst, including the split window algorithm (swa), sebal method, and single-channel algorithm (sca). after computing surface temperatures, spatial relationships between various independent natural and human factors and the calculated temperatures can be analyzed using geographic information system (gis) capabilities and statistical methods. over the last two decades, application of statistical and probabilistic models within the gis framework to assess the impacts of various factors on specific phenomena has gained considerable attention.the primary aim of this study was to extract trends in lst changes from 6 landsat 8 images taken from the eastern part of qazvin province during both summer and winter in 2016, 2019, and 2021,. the secondary objective was to investigate the relationship between changes in lst treated as a dependent variable and several topographical and human factors of the study area considered as independent variables across the two seasons. ultimately, the goal was to analyze the impacts of different classes of independent variables on the trends in surface temperature during the selected time period using the frequency ratio (fr) regression model. materials & methodsthe eastern part of qazvin province was chosen as the study area for this research. this region is significant for environmental assessments due to the presence of several industrial zones, tehran-karaj-qazvin freeway, which connects a large portion of eastern and central iran to the western provinces, and shahid rajaee power plant located along this route. to assess the trend in temperature changes, landsat 8 satellite images from the winters and summers of 2016, 2019, and 2021 were utilized. after downloading the images from the website of united states geological survey, initial preprocessing steps were conducted using envi software. these steps included radiance and atmospheric calibration to correct for sensor and atmospheric errors, as well as cropping the images to focus on the study area. following the calculation of surface temperature from the satellite images, the results were validated using the sebal method. the discrepancies were found to be less than 1°c. additionally, the calculated temperatures were compared with hourly air temperature data from qazvin meteorological stations recorded during the overpass of the landsat satellite, with differences remaining under 1.5°c. the final lst maps produced in envi were then imported into arcgis software for further classification, comparison, and spatial-temporal evaluation.to achieve the second objective, various data layers related to natural and human factors—including a digital elevation model, slope, aspect, land cover map, and road network—were compiled. the relationship between lst (the dependent variable) and these variables (independent variables) was analyzed using the fr regression statistical model. the fr method is a widely used data mining approach for modeling and forecasting various natural hazards and urban trends. research findingsthe analysis of lst trends indicated that the northern part of the study area had generally experienced lower average temperatures due to its higher latitude and elevation. in contrast, the southern part had a higher average temperature attributed to its lower latitude, reduced altitude, and flatter terrain. notably, while winter 2019 recorded lower temperatures than the other two years, the summer of 2019 exhibited higher minimum and maximum temperatures compared to the preceding years. a significant portion of the region had recorded temperatures ranging from 30 to 45°c. overall, the temperature calculations for the summer across the three years suggested a warming trend from 2016 to 2021. when evaluating the impacts of various factors on surface temperature, the slope direction layer demonstrated a greater influence than either the elevation or slope layers. generally, northern slopes (north, northwest, and northeast) exhibited higher frequency ratios during summer due to reduced sunlight exposure in winter. conversely, southern and southeastern slopes showed higher frs in summer compared to winter, reflecting increased solar radiation. by extracting relationships through the fr method, valuable insights into the influence of independent variables on regional temperature could be gleaned. in summer, barren lands were associated with a very high fr for the hottest temperature class, while this ratio was significantly lower in winter. agricultural and grassland areas also exhibited considerable heat during summer compared to winter. factors like soil type, terrain flatness, direct solar radiation exposure, and aspect were crucial contributors to the elevated temperatures in these regions. additionally, the impact of industrial activities on surface temperature increases was significant across all three years, affecting both winter and summer seasons. in each season during the selected timeframe, industrial fr values were consistently higher at elevated temperatures. areas designated as industrial zones and towns had recorded higher temperatures than surrounding urban areas. furthermore, the influence of vehicle traffic on roadways was particularly pronounced in winter, with barren lands adjacent to roads contributing notably to ambient air warming.the fr calculation process assessed the occurrence of a phenomenon as a dependent variable in relation to independent variable classes. this evaluation could assist in predicting natural events, such as floods, earthquakes, landslides, and other phenomena. discussion of results & conclusionthis study aimed to analyze the land surface temperature (lst) of the eastern part of qazvin province in two stages. in the first stage, surface temperature was estimated using two widely used calculation methods and the rate of temperature fluctuations during the selected time period was assessed. the results indicated a strong correlation between the two methods of estimating surface temperature, as well as a close alignment with the hourly air temperatures recorded by meteorological stations in qazvin. this consistency could be attributed to the incorporation of key parameters, such as the emissivity coefficient, spectral radiation, and brightness temperature from thermal bands.a detailed statistical evaluation of lst classes revealed a clear increase in temperature from 2016 to 2021.in addition to identifying temperature trends, it was also crucial to examine the factors influencing these changes. in the second stage, the effects of various layers—such as elevation, slope, aspect, land cover, and proximity to road networks—were evaluated using the frequency ratio (fr) method. the results demonstrated significant impacts from both human and natural variables on surface temperature fluctuations. by extracting these relationships, valuable insights were gained regarding the influence of independent variables on the dependent variable—surface temperature. these findings can be applied in various fields, including urban planning, environmental management, and crisis response. finally, for more accurate calculations and analyses of surface temperature, particularly in urban areas, the use of high-resolution data and images is recommended to effectively assess different urban land uses.