Sea Level Fusion of Satellite Altimetry and Tide Gauge Data by Deep Learning in the Mediterranean Sea

Satellite altimetry and tide gauges are the two main techniques used to measure sea level. Due limitations of satellite altimetry, a high-quality unified level model from coast open ocean has traditionally been difficult achieve. This study proposes fusion approach gauge data based on deep belief network (DBN) method. Taking Mediterranean Sea as case area, progressive three-step experiment was designed compare fused anomalies DBN method with those inverse distance weighted (IDW) method, kriging (KRG) curvature continuous splines in tension (CCS) for different cases. The results show that precision varies methods input measurements. is better than other three most schemes reduced by approximately 20% when limited along-track in-situ used. In addition, distribution large effect but less impact model. Furthermore, spatial resolution 0.25° × generated contain more information others, which means can be applied feasible robust way fuse these kinds levels.