A study of the relationship among temperatures of surface features and its application in remote sensing study of Lut desert

Abstract

Physical characteristics of different features in desert is a reflection of severe thermal and climatic conditions. In this paper, diurnal surface temperature patterns of important surface features in Lut Desert were studied and the relationship among different surfaces analyzed. Diurnal trend in surface temperature of surface types, marl, dark sand, light sand, salt-affected soil, soul at 10 cm depth, as well as dry and wet air temperature within 15 days were recorded in 2 hour intervals in the margin of Lut yardangs while correlations among these surface features and its significance level were investigated. The knowledge of diurnal temperature pattern and calculation of correlation among features can lead us to the understanding of the behavioral pattern and the trend of surfaces. The knowledge of behavioral pattern of correlation coefficients during different hours of the day provides the researchers the ability to produce optimal models of thermal characteristics and to predict them. Additionally, thermal data of different sensor systems along with their capability in the study of surface features in Lut Desert was evaluated. The results of thermal and inter-feature correlation analysis reveal a similar trend (at 8 AM and 4 PM) among various feature types. The correlation coefficient of different surfaces shows the highest value at 6 AM and the lowest at 8 PM. The correlation coefficient between sand and marl with grey sand and soil shows a non-linear trend at pre-noon hours while a linear and similar trend in the afternoon. Generally speaking, the correlation coefficient between surfaces at pre-noon hours as compared to afternoon hours is completely non-linear and does not reveal any specific pattern. It seems that TIR data acquired at 10 AM is the best choice in the study of Lut Desert and hence ETM + and TM data can be used.