L-band rough sea surface radiation brightness temperature simulation method based on machine learning

Abstract

The invention provides an L-band rough sea surface radiation brightness temperature simulation method based on machine learning, and the method comprises the steps of building a matching data set based on multi-source satellite observation data and reanalysis data; establishing a low-wind-speed sea surface radiation brightness temperature model based on a deep neural network model and a high-wind-speed sea surface radiation brightness temperature model based on a small sample learning method; determining different air-sea parameter combinations possibly influencing the salinity inversion precision by using a cross validation method; and inputting field observation data into the final radiation brightness temperature model to obtain a radiation brightness temperature result. The invention has the advantages of high calculation speed and high precision; a machine learning method is introduced into L-band rough sea surface radiation brightness temperature simulation, and the sea surface salinity inversion precision is improved.

Description

technical field [0001] The invention belongs to the technical field of microwave remote sensing, and specifically relates to a method for simulating brightness temperature of L-band rough sea surface radiation based on machine learning. Background technique [0002] Sea surface salinity plays an irreplaceable role in the study of global climate change and its regional responses, including ocean thermohaline circulation and heat transport, global ocean precipitation estimation, ocean mixing process, land-sea interaction, global water cycle monitoring, and climate forecasting. High-precision sea surface salinity remote sensing is a high-tech with broad development prospects, and it is one of the frontiers of oceanographic research. Satellite microwave remote sensing technology has become the most important means of global intensive detection of sea surface salinity due to its advantages of wide coverage, high spatial resolution and continuous observation. [0003] The real se...

AI Technical Summary

Problems solved by technology

[0004] At present, although the theoretical model of the L-band rough sea surface radiation brightness temperature has a strong generalizability, due to the complexity of the theoretical model modeling and the lack of unified standards for some key parameters of the model (such as the critical wave number of the dual-scale model), the When modeling, it cannot fully consider various possible influencing factors such as sea surface roughness, white cap, sea temperature difference, rainfall, etc.

Images