An Adaptive Reduced Method for Wind Direction Retrieval from X-Band Marine Radar Images

Abstract

An adaptive reduced method (ARM) for retrieving wind direction from X-band marine radar images is presented in this paper. The new algorithm design an adaptive reduced operator which can expand the scope of the radar image resolution to adapt the wind streak scale by adaptively reduce radar image size. Different from previous algorithms that construct an adaptive judgment algorithm to obtain the optimum reduction rate, by calculating the stability factor based on pixels gradient histogram statistics to determine the reduced image resolution adapt to the wind streak scale. Wind direction is determined by the image pixels gradient direction of optimal reduction rate and the wind streak feature. The algorithm has been tested using in-situ X-band radar image sequences and shore-based anemometer data collected on southeast coast of China. It is shown that the coefficient of correlation is 0.99 between retrieved from adaptive reduced method and reference wind direction, the standard deviation is 7.62° and the deviation is −1.04°. Compared with the previous LGM algorithm, the accuracy of retrieved wind direction is improved by 58.3% relative to anemometer data, which show that the proposed method is reasonable.