Relationships between the three components of air temperature and remotely sensed land surface temperature of agricultural areas in Morocco

ABSTRACT

In this article, correlations to derive minimum (T_min), mean (T_mean), and maximum (T_max) air temperatures from land surface temperature (T_s) in Morocco are studied. Land surface temperature was derived from 349 decadal cloud-free images acquired by the Advanced Very High Resolution Radiometer sensor, onboard the National Oceanic and Atmospheric Administration’s satellites. T_s was first compared with T_min, T_mean, and T_max, at the level of nine meteorological stations between 1995 and 2012. This first step showed the existence of good linear correlations between T_s and air temperatures. The coefficient of determination, R², varied between 0.59 and 0.79 for T_s versus T_min, between 0.60 and 0.76 for T_s versus T_max, and between 0.67 and 0.79 for T_s versus T_mean. The root mean square error (RMSE) varied between 2.4°C and 3.9°C for T_s versus T_min, between 2.5°C and 4.6°C for T_s versus T_max, and between 2.2°C and 3.8°C for T_s versus T_mean. The mean absolute error (MAE) varied between 2°C and 3.1°C for T_s versus T_min, between 1.9°C and 3.6°C for T_s versus T_max, and between 1.8°C and 3°C for T_s versus T_mean. Second, T_s was compared with T_min, T_max, and T_mean gridded at the level of agricultural areas of 46 provinces in Morocco. For 42 provinces, the mean value of R² was 0.71 for both T_min versus T_s and T_max versus T_s, and 0.76 for T_mean versus T_s. The mean values of the RMSE and the MAE were 3.1°C and 2.4°C for T_s versus T_min, 3.6°C and 2.8°C for T_max versus T_s, and 3°C and 2.3°C for T_mean versus T_s, respectively. Finally, the accuracy of the regression models between air and surface temperatures was tested using a k-fold cross-validation method that showed high stability of these models. The relationships obtained in this work could be very useful for further monitoring and modelling agriculture and meteorological parameters.

Acknowledgements

Satellites images used in this article were provided by the Flemish institute for technological Research in Belgium. The authors thank all those who contributed to the processing of these images. They are also indebted to the staff of the National Meteorology Direction of Morocco for providing climatic data and for their fruitful collaboration.

Disclosure statement

No potential conflict of interest was reported by the authors.