Earth and Space Science最新文献

In storm surge (SS) simulation, data-driven methods can establish the relationship between predictor variables and the predictand, enabling long-term SS level reconstructions. Here, using the U.S. East Coast as an example, we explored the capabilities of four machine learning algorithms, namely Artificial Neural Networks (ANN), Long Short-Term Memory (LSTM), Light Gradient Boosting Machine (LightGBM), and Extreme Gradient Boosting (XGBoost) in reconstructing hourly SS levels from 1979 to 2018 under an all-site modeling framework. Four atmospheric parameters, time index, and tide gauge coordinates from 51 tide gauges are used as predictors. The model performance was evaluated at both the tide gauge and coastal scales. Results indicate that LightGBM and XGBoost models outperform ANN and LSTM in SS reconstructions, with XGBoost showing better overall performance, especially for extreme SSs and historical extreme events. XGBoost can capture the temporal evolution of SSs with higher accuracy, producing reconstructions comparable to observations under the all-site modeling framework. The model interpretability analysis focusing on XGBoost reveals that the spatial distribution of feature importance varies for each predictor. Mean sea level pressure and the 10 m eastward wind component are the two most important predictors, followed by time index, latitude, and longitude under the all-site modeling framework and selected stations. These results indicate that data-driven models under this framework have the potential to capture region-specific and physically reasonable relationships between SS levels and atmospheric drivers.

Quality filtering of satellite XCO₂ retrievals is essential for every downstream science application, yet the community still grapples with the trade-off between retaining data availability and suppressing biases. For the Orbiting Carbon Observatory-2 (OCO-2) record in particular, the long-standing binary quality flag targets the single use case of global carbon flux inversion and is often too restrictive for small spatial scale analysis like quantifying emissions from coal-fired power plants. To address the need for flexible quality filtering we introduce a data-driven, ternary (three-state) quality flag constructed from the agreement of three independent sub-filters: two Random Forest classifiers trained on distinct “truth proxy” data sets and a third sub-filter based on the uncertainty estimate from a non-linear machine learning bias correction and the operational uncertainty product. We utilized a Bayesian multi-objective optimization to tune the sub-filters, balancing the competing goals of maximizing data throughput with minimizing error variance and retrieval uncertainty. The proposed ternary quality flag shows an improved reduction in root mean square error (RMSE) (22% for land and 53% for ocean) over the operational flag. This reduction is due in part to an improved ability to remove observations affected by 3D cloud biases. The flexible ternary flag can optionally increase data availability by 21% over land and 18% over ocean but is still competitive with the RMSE of the operational product. The proposed filter addresses the diverse needs of the science community and is generalizable to greenhouse gas monitoring missions such as GOSAT, CO2M and Carbon-I.

Modeling sea ice microwave emissivities at large scales presents challenges, due to complex interactions between the microwave signal and the sea ice environment. For the preparation of the Copernicus Imaging Microwave Radiometer mission (CIMR) that focusses on the monitoring of polar regions, a pragmatic parameterization of the sea ice emissivity over the Arctic in winter is proposed, providing consistent emissivity parameterizations between 1.4 and 36 GHz, for both orthogonal polarizations. Satellite-derived microwave emissivities are calculated from the Advanced Microwave Scanning Radiometer 2, Soil Moisture Active Passive, and Soil Moisture Ocean Salinity observations, subtracting the atmospheric contributions and the surface temperature modulation using ERA5 meteorological reanalysis. The resulting Arctic sea ice emissivities are analyzed, alongside sea ice geophysical parameters from neXtSIM model outputs and ERA5, to identify the variables for the emissivity parameterization. Sea Ice Thickness emerges as a crucial factor, particularly at 18 and 36 GHz. A training database of coincident satellite-derived emissivities and geophysical parameters is set up, to develop a Neural Network parameterization of the emissivities based on the geophysical parameters. This pragmatic methodology establishes a direct link between calculated emissivities and physical sea ice properties, eliminating the need for a priori assumptions. Promising emissivity results are obtained, with Root Mean Square Error below $��\sim �$0.03 for most channels, and reaching $��\sim �$0.04 at 36 GHz. Part of the error is expected to come from uncertainties in the input geophysical parameters. The emissivity frequency dependence is checked, and the emissivity angular variation of the 1.4 GHz is calculated from SMOS-derived emissivities.

The abundant nighttime (twilight) spectra from October 2024 are used for characterization of on-orbit instrument performance of the Tropospheric Emissions: Monitoring of Pollution (TEMPO, a geostationary imaging spectrometer). We select 250 lightning flashes and measure the flash sizes and flash positions in 6 spectral domains spread across the two TEMPO detectors. Some of the flash sizes come close to ≈1 pixel ground-based estimates of the instantaneous field of view in the North-South direction. On average, spectral measurements across the two detectors are co-registered to better than 0.15 pixels, thus also matching the pre-flight assessments.

Volcanic eruptions cause large-scale topographic changes, through the emplacement of lava flows and lava domes, the formation of craters and calderas, and thick ash and pyroclastic deposits. Here we analyze the TanDEM-X Digital Change Map (DCM), which compares the DEM produced during 2010–2015 with satellite acquisitions collected in 2016–2022. The DCM covers 159 eruptions at 103 volcanoes; the data was good quality at 44 of these but not useable at 28. Topographic changes associated with volcanic activity was visible at 58 volcanoes including lava flows, domes, intrusions, pyroclastic flows, lahars, tephra fall, crater formation and landslides. We analyze five case studies in detail: Sierra Negra, Galápagos; Erta Ale, Ethiopia; Sangay, Ecuador; Ebeko, Russia; and Nabro, Eritrea. Our measurements of the lava flows at Sierra Negra and Nabro and crater formation at Ebeko agree to within 15% of previous measurements, confirming the accuracy of the TanDEM-X DCM in volcanic areas. At Erta Ale, we find maximum lava thickness of >40 m, greatly exceeding previous field-based estimates (<2.5 m); consequently, our total volume estimate is an order of magnitude higher. At Sangay, the patterns of height change are consistent with local reports, but our measurements have high uncertainties due to the prevalence of vegetative noise and steep topography. Overall, we demonstrate that the TanDEM-X DCM can measure topographic changes at volcanoes, and in many cases allows us to make new measurements. Finally, we discuss the lessons learned from the TanDEM-X DCM for planning future satellite missions, including the upcoming European Space Agency Harmony Mission.

Global climate change is significantly impacting dust emission patterns in arid and semi-arid regions, posing challenges to environmental quality and human health. However, the impacts of future climate change on dust emissions in China remain insufficiently understood. This study employed the Weather Research and Forecasting coupled with Chemistry (WRF-Chem) model to project future dust emissions in China under two climate scenarios (SSP2-4.5 and SSP5-8.5) for the years 2030, 2060, and 2090. Results indicated that in the near term (2030 and 2060), dust emissions were projected to be higher under the high-emission SSP5-8.5 scenario compared to the moderate-emission SSP2-4.5 scenario. This suggests that increased greenhouse gas concentrations and associated climatic changes may enhance conditions favorable for dust generation, such as elevated temperatures and reduced soil moisture. By 2090, however, this trend may reverse, with SSP2-4.5 exhibiting higher dust emissions than SSP5-8.5. This reversal highlights the complex, non-linear interactions between long-term climate variables and dust emission processes, potentially due to changes in precipitation patterns, atmospheric circulation, and vegetation cover. The spatial distribution of dust emissions consistently remains concentrated in northwestern China and southern Mongolia across all scenarios and time periods, emphasizing the persistent role of major dust source regions like the Taklamakan Desert and the Gobi Desert. These findings underscore the need for targeted mitigation and adaptation strategies to manage the environmental and health impacts associated with dust emissions in the context of climate change.

Submesoscale dynamics can induce significant vertical fluxes of phytoplankton, nutrients, and carbon, resulting in biological and climatological impacts such as enhanced phytoplankton production, phytoplankton community shifts, and carbon export. However, resolving these dynamics is challenging due to their rapid evolution (hours to days) and small spatial scales (1–10 km) of variability. The Modular Aerial Sensing System (MASS), an airborne instrument package measuring concurrent ocean dynamics and hyperspectral ocean color, provides a powerful tool to study the influence of submesoscale dynamics on phytoplankton and carbon. In this study, we present the first airborne observations pairing snapshots of sub-kilometer ocean velocities and their derivatives (i.e., vorticity, divergence, and strain) with concurrent ocean color and sea surface temperature. We developed airborne proxies of chlorophyll-a and particulate organic carbon, which explained about 66.2% and 56.2% of in situ variability without atmospheric correction, suggesting that MASS can capture phytoplankton variability. We also explored relationships between concurrent vorticity, divergence, strain, sea surface temperature, chlorophyll-a, and hyperspectral variables to illuminate the submesoscale processes that alter phytoplankton distributions. This study demonstrates the value of merging bio-optical and physical airborne remote sensing data to better understand the influence of submesoscale dynamics on oceanic ecosystems and organic carbon. We highlight the potential for suborbital remote sensing for studying processes that impact phytoplankton ecosystems and carbon transport without the spatiotemporal aliasing affecting in situ sensors.

An accurate geoid has important consequences for many fields such as engineering applications, underground resource exploration, geophysical surveys, etc. Its precise determination relies on two key data sets: gravity measurements and high-resolution elevation data, both of which are critical for achieving reliable results. In particular, accurate elevation data is indispensable for geoid modeling, as it is required for various computational steps, including the prediction of the free-air gravity anomalies, terrain corrections, and the calculation of complete Bouguer gravity anomalies. In the absence of accurate regional elevation data, the digital elevation model (DEM) generated by the Shuttle Radar Topography Mission (SRTM) is commonly used as a reliable alternative. Additionally, researchers from Japan and the United States have released a new DEM generated from the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), which provides an alternative to the widely used SRTM DEM. This study explores the consequence of the ASTER DEM on estimating mean free-air gravity anomalies in geoid determination, focusing on the Colorado experiment area, which is characterized by mountainous and rugged terrain. Numerical results indicate that the ASTER DEM yields less favorable statistics compared to the SRTM DEM in terms of height accuracy. The use of ASTER DEM introduces discrepancies (compared to SRTM DEM) ranging from −2 to 4 mGal in the interpolation of free-air gravity anomalies. Furthermore, it is demonstrated that the geoid differences resulting from the use of ASTER DEM are within a few centimeters, remaining below the accuracy level of external GNSS-leveling data.

We present a revision and update to the high-precision relocated seismicity catalog presented by Matoza et al. (2021, https://doi.org/10.1029/2020ea001253) for the Island of Hawai'i from 1986 to 2018. The starting catalog of hypocenters (input data), on which the study by Matoza et al. (2021, https://doi.org/10.1029/2020ea001253) was based, contained an inconsistent depth datum for events before and after 00:00 UT, 29 December 2017. Here we present a recomputed version of the catalog using a consistent reference depth. We corrected the starting catalog to a common depth datum (all events now use the model depth reference datum) and re-ran the entire workflow as described in the paper by Matoza et al. (2021, https://doi.org/10.1029/2020ea001253). This included pairing, cross-correlating, and relocating all seismic events again based on the updated starting catalog. We consider 347,446 events representing 32 years of seismicity on and around the island from 1986 to 2018. We now successfully relocate 299,966 (86%) events using ∼2.53 billion differential times (P and S) from ∼194 million similar-event pairs, derived from cross-correlations between ∼887 million event pairs total, a significant increase from our original analysis. The resolution of fine-scale seismicity features is improved and the median depth of shallow events (<5 km) under Kaluapele (Kīlauea summit caldera) in 2018 is shifted 926 m deeper as a result of the change. The interpretations and other major conclusions in the paper by Matoza et al. (2021, https://doi.org/10.1029/2020ea001253) are unchanged.