Image-To-Image Training for Spatially Seamless Air Temperature Estimation with Satellite Images and Station Data

Abstract

Air temperature at approximately 2 m above the ground (Ta) is one of the most important environmental and biophysical parameters to study various earth surface processes. Ta measured from meteorological stations is inadequate to study its spatio-Temporal patterns since the stations are unevenly and sparsely distributed. Satellite-derived land surface temperature (LST) provides global coverage, and is generally utilized to estimate Ta due to the close relationship between LST and Ta. However, LST products are sensitive to cloud contamination, resulting in missing values in LST and leading to the estimated Ta being spatially incomplete.To solve the missing data problem, we propose a deep learning method to estimate spatially seamless Ta from LST that contains missing values. Experimental results on 5-year data of mainland China illustrate that the image-To-image training strategy alleviates the missing data problem and fills the gaps in LST implicitly. Plus, the strong linear relationships between observed daily mean Ta (Tmean), daily minimum Ta (Tmin), and daily maximum Ta (Tmax) make the estimation of Tmean, Tmin, and Tmax simultaneously possible. For mainland China, the proposed method achieves results with R2 of 0.962, 0.953, 0.944, mean absolute error (MAE) of 1.793 C, 2.143 C, and 2.125 C, and root-mean-square error (RMSE) of 2.376 C, 2.808 C, and 2.823 C for Tmean, Tmin, and Tmax, respectively. Our study provides anewparadigm for estimating spatially seamless ground-level parameters from satellite products. Code and more results are available at https://github.com/cvvsu/LSTa.