Hongming Yao

Estuarine waters exhibit strong spatiotemporal variability in the partial pressure of CO₂ (pCO₂ ), driven by complex land–ocean interactions. Single-sensor satellite observations struggle to simultaneously resolve the fine spatial structures and rapid temporal changes required for reliable pCO₂ mapping, making multi-sensor spatiotemporal fusion (STF) a promising solution. However, a systematic assessment of whether and how STF benefits downstream pCO₂ retrieval is lacking. This study generated fused reflectance products over the Pearl River Estuary by integrating gap-filled MODIS and Landsat 8/9 observations using seven representative STF models spanning weighting, unmixing, hybrid, and deep-learning paradigms. These products were integrated into a semi-mechanistic retrieval framework that separately resolves the mixing-controlled and biologically driven components of pCO₂ (pCO₂ -Bio). Results reveal a consistent trade-off between radiometric accuracy and local texture fidelity among fusion models. More importantly, the benefit of fusion for final pCO₂ accuracy was governed primarily by the preservation of spatial gradients, frontal structures, and fine-scale heterogeneity—as quantified by texture-sensitive metrics like Multi-Scale Local Binary Patterns—rather than by fusion-stage numerical error (e.g., RMSE). The biologically driven pCO₂ -Bio component was substantially more sensitive to upstream retrieval uncertainties than the mixing-controlled component pCO₂-Mix, with reflectance bands Rrs482, Rrs561, and sea surface temperature (SST) emerging as dominant uncertainty sources in the retrieval chain. By shifting the evaluation paradigm from pixel-level radiometric precision to application-driven texture fidelity and by pinpointing the dominant uncertainty pathways, this work establishes a practical framework for generating reliable high-resolution pCO₂ products in complex coastal waters, thereby laying the foundation for more accurate estuarine carbon flux quantification and integration into global carbon assessments.

In situ pH sensing is crucial for the real-time monitoring of ocean acidification and investigations into the marine carbon cycle. Although ion sensitive field-effect transistor (ISFET) has been proven suitable for marine pH monitoring, its supply and implementation remain challenging. An underwater pH sensor for environmental analysis (uSEA-pH) based on ISFET was developed herein, incorporating a modified commercial laboratory pH probe through engineering design. Laboratory characterization demonstrated that uSEA-pH exhibited a Nernstian response (slope -57.60 ± 1.05 mV/pH, R² > 0.999), rapid response time (∼7 s), and low measurement uncertainty (in situ deployments, buoy-mounted uSEA-pH reliably recorded tidal-driven pH fluctuations in Dapeng Bay (27 days) and Xiamen Bay (7 days), generating over 2.3 million field measurements. This study presents a viable, robust, and high-resolution approach for continuous pH monitoring in estuarine and coastal areas.

Estuaries are critical for land‐ocean carbon exchange, but coupling mechanisms between air–sea CO ₂ fluxes (FCO ₂ ) and phytoplankton gross primary productivity (GPP) remain poorly understood. This study used high‐frequency underway monitoring in the Lingdingyang Estuary to resolve spatiotemporal interplays between FCO ₂ and GPP. Annual mean FCO ₂ was 20.29 ± 23.34 mol C m ⁻² yr ⁻¹ , with flooding season (82.97 ± 80.49 mmol C m ⁻² d ⁻¹ ) an order of magnitude higher than dry season. Gross primary productivity averaged 2.23 ± 2.07 mol C m ⁻² yr ⁻¹ , increasing significantly during flooding. The results revealed a distinct “source‐to‐sink” FCO ₂ gradient, with a 116% reduction over ∼40 km, primarily driven by phytoplankton activity. Biological processes explained 30–50% of FCO ₂ variability. While net autotrophy in the mid‐estuary reduced FCO ₂ by 48.6 mmol C m ⁻² d ⁻¹ during flooding, heterotrophic activity downstream offset 40–60% of GPP‐driven uptakes. This study quantifies how urban estuary oscillate between carbon source and sink states, providing parameters for blue carbon frameworks and demonstrating that eutrophication‐driven loads reduce overall carbon sequestration efficiency through enhanced heterotrophic activities.

As coastal areas become more vulnerable to climatic impacts, the need for understanding estuarine carbon budgets with sufficient spatiotemporal resolution arises. Under various hydrologic extremes ranging from drought to hurricane-induced flooding, a mass balance model has been constructed for carbon fluxes and their variabilities in four estuaries along the northwestern Gulf of Mexico (nwGOM) coast over a four-year period (2014 – 2018). Loading of total organic carbon (TOC) and dissolved inorganic carbon (DIC) to estuaries include riverine discharge and lateral exchange from tidal wetland. The lateral exchanges of TOC and DIC reach 4.5 ± 5.7 and 8.9 ± 1.4 mol·C·m-2·yr-1, accounting for 86.5% and 62.7% of total TOC and DIC inputs into these estuaries, respectively. A relatively high regional CO2 efflux (4.0 ± 0.7 mol·C·m-2·yr-1) has been found, which is two times of the average value in North American coastal estuaries based on a previous study. Meanwhile, oceanic export is the major pathway for TOC (5.6 ± 1.7 mol·C·m-2·yr-1, 81.2% of total) and DIC (9.9 ± 2.9 mol·C·m-2·yr-1, 69.7% of total) loss. Yet uncertainties remain due to a varying extent of remineralization coastwide. In addition, the carbon budget exhibits high variability in response to the hydrologic changes. For example, storm or hurricane induced flooding can elevate CO2 efflux by 2 – 10 times in short periods of time. Flood following a drought state also increases lateral TOC exchange (from -3.5 ± 4.7 to 67.8 ± 17.6 mmol·C·m-2·d-1) but decreases lateral DIC exchange (from 28.9 ± 3.5 to -7.1 ± 7.6 mmol·C·m-2·d-1). The large variability of carbon budgets highlights the importance of high-resolution spatiotemporal coverage under different hydrologic conditions, as well as the significance of carbon contribution from tidal wetland on coastal carbon cycling.