Introduction
Land surface temperature (LST) is a key physical parameter of land surface processes, at local and global scales, which is a combination of all results from the land surface and the energy flow between the surface and the atmosphere. Earth surface temperature is an important indicator related to climatic, meteorological, hydrological and environmental phenomena and processes. Currently, data from meteorological stations are the most important decision-making reference in this case (Baidy et al., 2013: 517). What is considered as a basic defect in monitoring the temperature of the earth's surface is the lack of sufficient meteorological stations to know the temperature values in places without stations. Considering the limited information, the need to use remote sensing technology with time conditions, along with the feature of continuity and data collection in wide ranges, can be very efficient (Kake Memi et al., 2019).
Soil moisture is the amount of water stored in soil particles and is affected by factors such as precipitation, temperature and other soil properties (Pandi et al., 2020: 1). Estimating soil moisture is of great value for weather forecasting, climate change monitoring, and flood monitoring (Beau et al., 2018). Remote sensing techniques provide tools for mapping soil moisture at large spatial and temporal scales. Remote sensing can be effectively used to estimate soil moisture because soil light reflectance and thermal emission are highly correlated with soil moisture (Acharya et al., 2022: 2).
Meteorological and hydrogeological studies of the Sentinel satellite series have been developed by the European Space Agency (ESA) to support the services of the European mission and the demands of the Copernicus program (Zarei et al., 2021: 3980); The first three Sentinel missions contribute to the understanding of the Earth system by detecting, monitoring, and evaluating changes in the ocean, troposphere, and land components (Ruskas et al., 2016: 1).
Methodology
The studied area is Nair city. This city covers 8% of the total area of Ardabil province with an area of 1495 square kilometers. It is located at 47 degrees 59 minutes east longitude and 38 degrees 2 minutes north latitude. This city of Nir is located 35 kilometers west of Ardabil between Sablan and Bezgosh mountains. In this study, Sentinel 2 satellite images have been prepared for the date of 12/7/2022, and then atmospheric and radiometric corrections were made in ENVI 5.6 software, and a part of the images was cut based on the study area. After the atmospheric and radiometric correction of the image, the land use classification map was prepared in eCognition 9 software using the nearest neighbor method of the object-oriented algorithm; There are two main steps in image base object classification. The first step is image segmentation, which divides the image into separate areas or objects (segments) based on a similar spectral and spatial pattern. After segmentation, the second step is to relate the image segments using spectral and shape statistics, texture parameters and topological information. To perform this method, the results of different scales were analyzed. Finally, considering the scale of 60 and the amount of softness and compression, respectively, 4 0.0 and 0.6 segmentation was done. After segmentation of the study area, they were classified into seven classes: irrigated agriculture, rainfed agriculture, residential area, water areas, snow cover, pasture and communication road; Then, the kappa coefficient and accuracy were estimated for validation. Sentinel 2 image was also used to calculate soil moisture using OPTRAM method. To perform this method, first NDVI and STR index were calculated and then soil moisture value was determined using these two indices.
In the continuation of the research, to calculate the temperature of the earth's surface, the daily temperature product of Sentinel 3 images related to the time of 7/11/2022 was prepared for the studied area; And in the SNAP software, the temperature of the earth's surface was estimated. The Sentinel 3 temperature product is an official level 2 product and has a spatial resolution of 1 km. It provides estimates of LST and some related parameters, e.g. LST uncertainty, normalized difference vegetation index (NDVI), vegetation type (biome), atmospheric column water vapor content (CWVC) and parameters related to LST retrieval. has After extracting maps of surface temperature and soil moisture using satellite images, the correlation between the dependent variable of humidity and the independent variable of temperature has been investigated using geographic weighted regression.
calculating
In this research, in the first step, the classification of land use in seven classes of irrigated agriculture, rainfed agriculture, pasture, residential area, water areas, snow cover and communication road has been done using the object-oriented method in eCognition software. The accuracy of this classification has an overall accuracy of 99% and a kappa coefficient of 98%, which has achieved acceptable results due to the use of Sentinel 2 images. One of the influencing factors on land suitability for different uses is soil surface temperature. In this study, the temperature products of Sentinel 3 images were used to check the soil surface temperature. One of the factors affected by climate and temperature is soil moisture; In this way, Sentinel 2 images and OPTRAM method have been used to estimate soil moisture. Examining the results shows that the temperature of the ground surface and the amount of soil moisture are completely dependent on the nature of the types of use. The highest average temperature related to pasture use is 43 degrees Celsius and the lowest average temperature related to snow and water use is 34 and 37 degrees, respectively. Examining the soil moisture of each land use shows that water areas have the highest average humidity and residential areas have the lowest average. Based on the results, there is a direct and inverse relationship between soil surface moisture, vegetation density and surface temperature. The presence of moisture on the surface of the soil and vegetation leads to a decrease in the temperature of the earth's surface. So that dry lands, or in other words, low humidity, as well as lands with low density vegetation, correspond to the areas that show high temperature in thermal images.