In this paper, the spatial and temporal coverage of satellite SST in the Arctic region is studied by using the SST data of polar orbit spaceborne microwave radiometer (Windsat, AMSR2, HY-2A RM, GMI) in 2016, and the accuracy of SST data is evaluated by using the measured data of Argo. The results show that, the space borne microwave radiometer SST retrievals coverage rate and effective coverage days in winter are lower than that in summer. When AMSR2, GMI, WindSat and HY-2A RM space borne microwave radiometer SST data are combined used, the SST coverage rate can be between 12%-15% in February, and the number of effective observation days is better than 26 days. The error of the space borne microwave radiometer SST data in the Arctic is larger than that of the global average. The accuracy of AMSR2 data is the best one, the accuracy of WindSat data is close to that of AMSR2. The RMSE of GMI SST is about 2 times larger than AMSR2, and the accuracy of HY-2A RM data is lower than that of any other space borne microwave radiometer.

Radiometers working in the infrared bands are easily affected by factors such as clouds and sea fog, which limits the spatial coverage of sea surface temperature (SST) data observed by remote sensing and affects the application of SST remote sensing data products. Data Interpolation Convolutional Auto-Encoder (DINCAE) is a data reconstruction method based on deep learning that extracts nonlinear relationships in data through a convolutional auto-encoder structure and reconstructs missing data using valid points available in the data. On the basis of DINCAE, we added the convolutional LSTM (ConvLSTM) to fully extract the time features in the data and improved the DINCAE method (T-DINCAE). The reconstruction error is calculated through cross validation and Argo buoy data. The results show that T-DINCAE has higher data reconstruction quality than DINCAE.

In this paper, based on the optimal interpolation (OI) algorithm, a study on SST data fusion was performed using satellite-derived SSTs from three passive-microwave (WindSat, ASMR2 and HY-2B SMR) and five infrared (AVHRR, MODIS onboard Terra and Aqua, VIIRS and HY-1C COCTS) radiometers to generate 12h/4km high-resolution SST data in Southeast Asia and its adjacent seas in 2020. The accuracy was assessed with the measured Argo buoy observations. Comparisons with the measurements from Argo buoys show that the RMSEs of the fusion SST data are 0.4010°C for nighttime and 0.4180°C for daytime. The deviation distributions during daytime and nighttime are consistent, which is mainly reflected in the negative deviations in the 9°N-9°S sea areas and the positive deviations in the other sea areas.

High temporal resolution sea surface temperature (SST) fusion data for the Arctic were generated using optimal interpolation method with a spatial resolution of 9 km and a temporal resolution of 12 h. The accuracy of the fusion data was assessed using Argo in-situ measurements. The root mean square error (RMSE) of the fusion SST data (a temporal resolution of 12 h) ranged from 0.49°C to 0.62°C. The fusion SST data was more accurate than the AVHRR OISST data. There was no significant difference between the fusion SST data and RSS MW-IR SST data. This fusion SST data has high spatial coverage, with high temporal resolution and high precision.

The accuracy of SST derived from passive-microwave (AMSR2, GMI, and WindSat) and infrared (AVHRR, MODIS onboard Terra and Aqua and VIIRS) radiometers in Southeast Asia and its adjacent seas is evaluated using Argo observations. The results show that the root mean square errors (RMSEs) of the differences between SST from passive-microwave radiometers (AMSR2, GMI, and WindSat) and Argo SST are 0.46°C, 0.42°C, and 0.50°C, respectively. The RMSEs of the SST differences between infrared radiometers (VIIRS, AVHRR, MODIS onboard Aqua and Terra) and Argo buoy are 0.54°C, 0.63°C, 0.65°C, and 0.64°C, respectively. In addition, the spatial coverage of the satellite SST products in Southeast Asia and its adjacent waters is calculated. The daily coverage of AMSR2 SST ranges from 44.51 % to 66.74%, with annual average coverage of 56.8%. The yearly average coverage of MODIS onboard Aqua and Terra and VIIRS SST are 21.79%, 22.91%, and 20.67%, respectively.

The Chinese HY-1C satellite was launched on September 7, 2018, equipped with the Chinese ocean color and temperature scanner (COCTS). The COCTS was designed for sea surface temperature (SST) and ocean color detection. In this paper, the accuracy of SST derived from COCTS onboard HY-1C in Southeast Asia and its adjacent seas is evaluated using Argo observations. And SST product from VIIRS onboard NPP is considered as a comparison. The bias, standard deviation (STD), and root mean square error (RMSE) of the SST difference between HY-1C and Argo data are 1.76°C, 1.60°C, and 2.38°C for daytime. The precision of HY-1C SST is significantly lower than that of VIIRS. HY-1C COCTS SST is corrected using VIIRS SST. The bias, STD, and RMSE of the SST difference between the corrected HY-1C and Argo data are 0.17°C, 1.21°C, and 1.22°C, which are better than the results before correction.

Due to the influence of clouds and other factors on radiometers operating in the optical and infrared bands, satellite data of sea surface temperature (SST) usually has large areas missing. This will have a severe impact on its application. In this study, Data Interpolation Empirical Orthogonal Function (DINEOF) method was applied to reconstruct the missing values of SST in a region in Southeast Asia. The DINEOF was used to the 2019 Moderate Resolution Imaging Spectroradiometer (MODIS) Aqua L3 SST data, and its accuracy was assessed by comparing Argo data, satellite data, and reconstructed data. And Argo data was used to correct the reconstructed data. The corrected Std and RMSE were 0.3992 and 0.3951, respectively.

Two linear regression algorithms were developed for the retrieval of summer and winter sea surface salinity (SSS) using the Moderate Resolution Imaging Spectroradiometer (MODIS) remote sensing reflectance (Rrs) products, together with in situ SSS measurements from the Bohai Sea. The proposed models were applied to MODIS monthly 4km Rrs data to study SSS variations of the Bohai Sea during 2003–2018. The coastal waters in Laizhou Bay, Bohai Bay, and Liaodong Bay were found to be fresher than in the Qinhuangdao sea area, Central Bohai Sea, and Bohai Strait. The SSS values in winter in the Bohai Sea were >31.00 psu and <31.50 psu as a whole. There is a rising trend of SSS in the coastal areas of the Bohai Sea, especially in the Liao River estuary, northern Weihe River estuary, northeast Luan River estuary and in the north part of ancient yellow river course.

The capacity of L-band microwave radiometers to observe salinity over coastal waters and produce satisfactory salinity fields remains limited, because of contamination from radio frequency interference (RFI). The objective of this study was to explore the potential of MODIS to derive sea surface salinity (SSS) for the coastal waters in the Bohai Sea. This study developed two linear regression algorithms for the retrieval of summer and winter SSS of coastal waters using the Moderate Resolution Imaging Spectroradiometer (MODIS) remote sensing reflectance (Rrs) products, together with in situ SSS measurements from the Bohai Sea. Evaluation using independent in situ measurements showed that the root mean square error and R2between the in situ SSS measurements and predicted SSS in summer were 0.20 psu and 0.76, and were 0.40 psu and 0.64 in winter, respectively.

Using the AVHRR OISST data from 1982 to 2019, the spatial and temporal characteristics of SST in the North Pacific Ocean were analyzed. The relationships between SST (Sea Surface Temperature) and SLP (Sea Level Pressure) as well as SSWF (Sea Surface Wind Field) were also discussed based on the ERA5 data. The study showed that the mean value of SST increased slowly at a rising rate of O.12°C/10yr during 1982–2019. EOF analysis showed that the dominant modes in the North Pacific were ENSO mode, consistent warming mode and dipole mode. There was a positive correlation of 0.58 between SLP and SST in the Aleutian low pressure area, while there was a negative correlation relationship from east Hawaiian Islands to the California seas. SST was mainly influenced by easterly winds in the area north of 36°N, while by southerly winds in the sea south of 36°N.

Triple collocation analysis is an effective method for assessing the error of observation systems with independent random error. However, this error independence is often violated, and systems show clear error correlations. Whereas many efforts have been devoted, the cause and determination of this error correlation are still challenging. In this work, we reveal that the error correlation is actually a signal observed by some systems, and taking this signal as the error correlation between systems will bias the error estimation results. Therefore, we define it as the representativeness signal instead of the representativeness error. Based on the multiscale signal approach and extended collocation analysis, we discuss the cause of this representativeness signal and its effect on sea surface salinity (SSS) data validation by synthesized experiments, satellite, buoy, and climatology products. The results suggest that the representativeness signal may vary with the different behaviors of observed variables. For SSS, the representativeness signal lies in satellite data with medium resolutions and similar representativeness. However, thein situand climatology data, which have different resolutions compared with satellites, do not show apparent representativeness signals with satellite data. The data collocation procedure also affects the representativeness signal, which decreases with increasing collocation intervals. The extended collocation analysis can estimate the error of SSS products in most cases but may not provide robust estimation when the collocation data pairs are limited.

Reef detection is of great significance in providing safe navigation route, developing and optimizing the marine conservation planning. This paper used the gradient vector flow active contour model, also known as the GVF snake model, applying SPOT-6 satellite remote sensing image to carry on the reef edge detection of Yin LiTan area, avoiding the complicated process that the profile outline need to be drew manually in the past. The experiment results show that the boundary detected by the GVF snake model is consistent with the in situ data from Island Placename Special Project in the year of 2013, and the model is suitble for reef edge detection.

Water depth is necessary for shipping security, port and marine engineering construction and planning utilization of coastal zone and island. Traditional bathymetry methods can provide accurate measurements of the water depth. But remote sensing has characteristics such as speediness, synchronization, large area, high resolution and so on, and it can solve many problems which are hard to overcome in traditional fathoming method. To fully utilize the existing resources of remote sensing images and excavate multi-temporal information effectively, the paper proposed the fusion model of multi-temporal remote sensing bathymetry inversion in Beidao Island based on fuzzy membership degree. By the help of decision fusion, both the mean relative error and the mean absolute error of the multi-temporal fusion result are lower than those of the single-temporal inversion results.