Recently, the research team led by Researcher He Xianqiang from our laboratory, along with collaborators, has published a research paper titled Retrieval of sea surface salinity from SMAP L-band radiometry under rainfall conditions in Remote Sensing of Environment. The first author of the paper is Associate Researcher Jin Xuchen from our laboratory, and the corresponding author is Researcher He Xianqiang of our laboratory. The co-authors include Academician Pan Delu, Researcher Bai Yan, Assistant Researcher Zhou Lizhang, Associate Researcher Li Teng, Senior Engineer Gong Fang and other researchers from our laboratory. This study marks a new breakthrough achieved by He Xianqiang’s research team in the field of ocean remote sensing under high sea state conditions.
Sea surface salinity (SSS) is a critical parameter characterizing ocean freshwater budgets and the global water cycle. Satellite L-band microwave radiometers are capable of large-scale sea surface salinity observation. Nevertheless, rainfall alters atmospheric radiative transfer and sea surface microwave emission properties simultaneously, leading to prominent negative salinity biases. Current operational products commonly exclude affected observations via rainfall flagging and quality control. While this measure stabilizes data products, it discards vital information required for investigating freshwater input from precipitation, storm evolution, and rapid responses of the upper ocean.
To address the above challenges, this study constructs a physical retrieval framework for SMAP sea surface salinity applicable under rainfall conditions. Firstly, CMORPH precipitation data are aggregated with Gaussian weighting matching the ~40 km SMAP observation footprint to obtain equivalent rainfall intensity. Next, atmospheric emission and attenuation induced by rainfall are corrected using a radiative transfer model. Meanwhile, both ring wave enhancement caused by raindrop impact and damping effects of rainfall turbulence on short gravity waves are incorporated into the sea surface roughness model. Finally, corrected specular brightness temperature is converted into sea surface salinity via the maximum likelihood method (Figure 1).
Figure 1 Flowchart of the SMAP sea surface salinity retrieval algorithm under rainfall conditions
Three independent datasets of near-surface salinity, namely drifter measurements, Salinity Snake observations and Wave Glider records, are adopted for model validation. Validation results show that uncorrected SMAP sea surface salinity generally presents negative biases ranging from 0.28 psu to 0.50 psu under rainy conditions. After rainfall correction, residual biases are limited within approximately ±0.16 psu, and the root mean square error (RMSE) is relatively reduced by 15%–55%. Taking the 2-hour matching window as an example, the bias of surface salinity validation declines from 0.493 psu to 0.069 psu, and the RMSE drops from 0.546 psu to 0.245 psu, demonstrating that the proposed method can significantly mitigate the systematic artificial freshening induced by rainfall.
For heavy rainfall cases, uncorrected salinity yields abnormally low values below 29 psu in the rainfall core area, while corrected salinity rises to around 31 psu and exhibits better consistency with footprint-scale freshening distribution predicted by rainfall impact models (Figure 2). Experiments with different precipitation datasets and footprint weighting schemes verify that the proposed method can robustly suppress rainfall-induced low-salinity biases. This study confirms that rainfall-affected L-band observations still contain valid salinity information, offering a novel approach to monitor near-surface salinity variations associated with ocean freshwater cycling and storm events.
Figure 2 Spatial comparison of rainfall rate and sea surface salinity before and after correction for a heavy rainfall case
References
- Jin, X., He, X.*, Bai, Y., Wang, Y., Zhou, L., Yu, S., Li, T., Gong, F., & Pan, D. (2026). Retrieval of sea surface salinity from SMAP L-band radiometry under rainfall conditions. Remote Sensing of Environment, 345, 115570.
Related Publications
-
Jin, X., He, X.*, Shanmugam, P., Bai, Y., Ying, J., Zhu, Q., Zhao, Y., & Pan D. (2026). Estimation of sea surface foam coverage and effective foam layer thickness from satellite microwave measurements. Remote Sensing of Environment, 334, 115176.
-
Jin, X., He, X.*, Shanmugam, P., Ying, J., Gong, F., Zhu, Q., & Pan, D. (2024). Modeling the influence of precipitation on L-band SMAP observations of ocean surfaces through machine learning approach. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 17, 10291-10305.
-
Jin, X., He, X.*, Wang, D., Ying, J., Gong, F., Zhu, Q., Zhou, C., & Pan, D. (2023). Impact of rain effects on L-band passive microwave satellite observations over the ocean. IEEE Transactions on Geoscience and Remote Sensing, 61, 4200316.
-
Jin, X., He, X.*, Bai, Y., Wang, D., Ying, J., Zhu, Q., Gong, F., Zhou, C., & Pan, D. (2023). A vector radiative transfer model for simulating the microwave emissivity of sea foam based on matrix operator method. Frontiers in Marine Science, 10, 1103843.
-
Jin, X., He, X.*, Shanmugam, P., Bai, Y., Gong, F., Yu, S., & Pan, D. (2021). Comprehensive vector radiative transfer model for estimating sea surface salinity from L-band microwave radiometry. IEEE Transactions on Geoscience and Remote Sensing, 59(6), 4888–4903.
