Analysis of Temperature Changes, Vegetation, Evapotranspiration, and Drought Based on Modeling of Landsat 5 and 8 Satellite Imagery and Meteorological Indicators

Remote sensing data can provide the latest information for studying land cover and its changes. Land cover detection helps in area management and decision-making. In addition, comparing images from a certain period can show the trend of changes in any region. The formation of human settlements is always based on natural factors such as water and soil, and villages, as the first form of human coexistence in a natural area, are influenced by numerous natural, economic, and other features. Vegetation is the first and most important producer of any ecosystem and is reflected in many factors; therefore, by studying the relationship between its changes and other factors such as temperature, evapotranspiration, and drought, the interaction of these factors can be understood. Monitoring changes in surface temperature due to seasonal and non-seasonal variations is essential because of their large impact on the human and natural environment. Remote sensing satellite imagery can be used to monitor surface temperature continuously, inexpensively, and rapidly.

The objective of this study was to investigate the spatial autocorrelation of surface temperature and evapotranspiration about vegetation in Maraveh Tappeh. Using data processing in the Google Earth Engine environment, the first part analyzes and evaluates a high-resolution seasonal dataset for each year over a period between 2000, 2015, 2010, and 2020. We present the analysis. Then, evapotranspiration was calculated using Modis data and the Torrent-White method, taking into account the plant growth coefficient, and compared with the evapotranspiration determined by the Sabal algorithm. In this context, drought prediction using the SPI index (one month, three months, six months, and twelve months) was also performed for the period 2015 to 2044. After calculating the changes, ArcGIS software was used to detect and discover patterns and trends in the spatial data. To obtain the best result and the lowest error rate, the method of quadratic exponential interpolation was used, and the mean square error RMS absolute value of MAE error was used to select the best interpolation method.

The results showed that the inverse distance method is the best method for estimating changes. The results of this study show that in spring, the highest percentage of NDVI does not correspond to the lowest temperature in space, in other words, the percentage of vegetation index is not inversely related to the temperature of the soil surface. In summer, the highest percentage of vegetation index is fully spatially matched concerning the space with the lowest ground surface temperature. In winter, the distribution of temperature patterns is completely different compared to other seasons because vegetation acts as a temperature moderator through the evapotranspiration mechanism.

Most areas with lower temperatures have denser vegetation. The correlation of vegetation has a value of -0.77, indicating a direct and inverse relationship between vegetation and temperature. The quantitative and qualitative changes of vegetation in the last 20 years have been extensive for the study area so the areas without vegetation have decreased somewhat and the rich vegetation has increased.