Multi‐layer perceptron neural network based algorithm for estimating precipitable water vapour from MODIS NIR data

Abstract

This Letter presents a multi‐layer perceptron neural network (MLP‐NN) based algorithm to quantitatively determine precipitable water vapour (PWV) directly from near infrared (NIR) radiance measured by the Moderate Resolution Imaging Spectroradiometer (MODIS). First, the background of the MLP‐NN based algorithm is discussed briefly. Then, the radiance of MODIS NIR channels simulated through a radiative transfer model with a set of input variables covering a broad range of surface reflectance and water vapour content are used to train MLP‐NN. Finally, PWV values derived by the MLP‐NN based algorithm are compared with radiosonde observations and a root mean squared error of 5.2 kg m⁻² is found from this comparison.

Acknowledgments

This work was funded by the National Natural Science Foundation of China under Projects 40471099 and 49890330, and the Major State Basic Research Project of China under grant 2002CB41250. The data used in this study were acquired as part of NASA's Earth Science Enterprise and archived and distributed by the Goddard Earth Sciences (GES), Data and Information Services Center (DISC) and the Distributed Active Archive Center (DAAC). The authors are grateful to the anonymous reviewers for their useful comments, which helped to improve the quality of this Letter greatly.