Earth and Space Science最新文献

Retrievals of ocean color (OC) properties from space are important for understanding the ocean ecosystem, the carbon cycle, and monitoring events such as harmful algal blooms (HABs). The recently launched U.S. National Aeronautics and Space Administration (NASA) Earth Venture Instrument, the geostationary Tropospheric Emissions: Monitoring of Pollution (TEMPO), provides a unique opportunity to examine diurnal variability in ocean ecology across coastal waters of North America and prepare for future hyperspectral geostationary OC missions. Although TEMPO does not match the spatial resolution or spectral coverage of planned coastal ocean sensors, such as NASA's Geosynchronous Littoral Imaging and Monitoring Radiometer or the U.S. National Oceanic and Atmospheric Administration Geostationary Extended Observations Ocean Color Instrument, it provides hourly observations at approximately 5 km over U.S. coastal regions and the Great Lakes. Here, we apply a newly developed atmospheric correction approach based on principal component analysis combined with machine learning (ML) to retrieve OC properties using TEMPO's hyperspectral measurements. Principal component coefficients derived from measured reflectances are used to train a neural network to estimate OC properties, including chlorophyll concentration, informed by collocated physically-based retrievals from MODIS, VIIRS, and Ocean and Land Color Instrument. This ML-based approach complements traditional radiative transfer retrievals, particularly under challenging conditions such as glint and moderate cloud coverage. This approach demonstrates the value of near-real-time OC products, with significant potential for monitoring HABs and transient oceanic phenomena.

The injection of CO₂ underground into reservoirs for carbon capture and underground storage (CCUS) is highly sensitive to heterogeneity and anisotropy. Although conventional geological modeling cannot take explicit account of heterogeneity or anisotropy or operate at a resolution that encompasses the small scales at which fluid flow is controlled, advanced fractal reservoir models (AFRMs) can be used. These AFRMs require fractal dimensions and anisotropy ratios for the target reservoir, and these values are unavailable. This paper describes the development, validation, and application of a software tool for measuring reservoir fractal dimensions. The code has been validated extensively using SynFrac data, recognizing four potential sources of systematic error, all of which can be corrected for. The resulting code has been used to measure the fractal dimension of a seismic data cube taken from the Chandon reservoir. The analysis reveals that the reservoir is multifractal with high heterogeneity (fractal dimension) at small scales and lower but still significant heterogeneity at larger scales. The fractal dimension can be calculated as a function of depth, providing a new type of log data that is not specific to a given well but rather specific to an area of seismic data.

The b-value from the Gutenberg-Richter law is a crucial parameter in the assessment of seismic hazard. Its temporal variations may also bring useful insights on the processes driving seismicity at depth, even if not yet fully understood. In this paper, we focus on the temporal evolution of the b-value in the Ubaye Region (French Western Alps) which was hit by seismic swarms (2003–2004) and complex sequences with several mainshocks (2012–2015). The swarm-like sequences show a common temporal behavior of b-value characterized by an increase and then a return to the initial level. The temporal b-value pattern for the mainshock-aftershock-like sequences is quite different. After a drop in the b-value that may follow the mainshock, the b-value increases above the background level before going back to it. Moreover, no precursory pattern can be identified before the mainshock. Fluid processes are recognized to play a crucial role in the driving mechanisms of these seismic sequences. Drawing parallel between swarms and aftershock sequences suggests that the b-value depicts fluid-processes in the Ubaye Region seismicity. We propose that b-value shows a complex behavior, with variations due to Coulomb stress-transfer from the mainshock and fluid-pressure processes. Therefore, even with a catalog made at the French national scale, the b-value variations may help to monitor the on-going processes at depth.

Interdisciplinarity is essential for addressing complex scientific problems that transcend disciplinary boundaries. Geology leverages methods from diverse domains to drive research innovation. However, quantitative evaluations of interdisciplinary connections between geology and other domains are lacking. Therefore, this study employed bibliometrics and natural language processing to assess the interdisciplinary trajectory of geology by analyzing temporal patterns in citation flows. First, a data set of geology-related publications and their cited references was collected from the Scopus database. Then, a semantic text classification approach, integrating sentence transformers and cosine similarity, was implemented to categorize cited references into eight scientific domains: mathematical and physical science, chemical science, life science, engineering and materials science, earth science, information science, management science, and health science. Longitudinal analysis of the distribution of references across these domains reveals trends in interdisciplinary collaboration over time. Finally, N-gram frequency analysis was performed on reference data associated with high-growth domains to identify specific influential techniques bridging disciplines. The results demonstrate a pronounced increase in interdisciplinarity between geology and information science, especially in applications of artificial intelligence, since 2016, with an average interdisciplinary potential of 0.1484. Key techniques driving this integration include artificial neural networks, logistic regression, support vector machines, etc. Additionally, the eight domains were expanded into 126 sub-disciplines to enable more detailed interdisciplinary analysis. Furthermore, three large language models were employed to verify the reliability of the adopted semantic analysis. The results suggest that our methodology provides robust approaches for quantifying interdisciplinary dynamics and is generalizable to other interdisciplinary fields.

Content assessment of research metrics plays a pivotal role in the evaluation of scientific productivity globally, especially in a selected field and region. Data from 28 Space-Science Journals spanning 2014–2023, from the Scopus-database, based on African publication output, citations, views-counts, and Field-Weighted-Citation-Impact (Field-Weighted Citation Impact (FWCI)) metrics were used. The results revealed that Africa contributes only 3.2% of the world publication volume in Space Science. From the African output, South-Africa leads with 40.9%, followed by Nigeria (14.3%) and Egypt (13.6%). These three countries contribute ≈70% of the African publication volume. For the citation metrics, Africa contributed 5.0% of the world volume. Publication in Journal of Advances in Space Research is more sought after by African Authors, while Astrophysics and Space Science journal recorded the highest African-to-world publication output percentage (11.3%). African authors show a preference for publishing in Journals with high percentile score and citation rates. Citation-wise, South-Africa accounted for 64% of the total volume from Africa. Only seven countries present citation metrics above 1% of the total volume. South Africa (46%), Morocco (10%), Egypt (9%), Namibia (7%), and Nigeria (7%) are the five countries with publication View counts of above 4,000. Only Ethiopia and South-Africa had FWCI above the world average, with values of 1.47 and 1.25 respectively. North Africa region dominated the appearance list of the 10 top countries in publication, citation, counts views and FWCI while Southern Africa leads in volume. The work further situates the uniqueness/global acceptance of the Scopus and Web-of-Science databases as tools for research publication assessment.

Tweek atmospherics are ELF/VLF pulse signals with frequency dispersion characteristics that originate from lightning discharges. Previous research has employed tweek atmospherics to examine long-term trends in the lower ionosphere; however, their utility in capturing diurnal-scale variations has been largely unexplored. Based on the machine learning method, we statistically study a massive data set of 48,395 first-order tweeks and obtain the diurnal variations of the nighttime lower ionosphere with a time resolution of 15 min. The variation amplitude of the mean reflection height ($��\Delta �h�$) in a single night could reach 7 km for the first-order tweeks, with an electron density variation ($��\Delta �N_e�$) of 2.5 cm⁻³. By comparison with the ionosonde observations from Wallops Island station and the incoherent scatter radar (ISR) observations from Millstone Hill station, we find that the correlation between the tweek-inferred lower ionosphere and the F2-layer electron density varies systematically with geomagnetic activity. In addition, the state of the tweek-derived lower ionosphere is also related to the E-region ionosphere (110–130 km), suggesting the presence of localized coupling process in this altitude range. Moreover, the comparison provides independent validation of our inversion technique with tweek atmospherics and confirms its potential to build a fully automated monitoring system for the nighttime lower ionosphere.

This work examines atmospheric effects on cosmic ray counts observed by a Water-Cherenkov detector at the Argentine Antarctic Marambio Station. We analyze the influence of ground-level barometric pressure and geopotential height at various pressure levels on daily particle rates, finding the strongest association at 100 hPa, linked to effective muon production. This relationship persists across low and high frequencies relative to the annual wave. Using barometric pressure and 100 hPa geopotential height, we developed a multiple linear regression model to describe atmospheric variations in cosmic ray flux, adjusted by meteorological seasons. By inverting the model, we estimate 100 hPa geopotential height from surface observations and validate against ERA5 reanalysis. The model performs best in spring, with reduced precision in other seasons. Further improvements in the signal-to-noise ratio could enhance model performance. Even with these considerations, this approach offers a practical and cost-effective method to track 100 hPa geopotential height variability in Antarctica through daily surface observations from Water-Cherenkov detectors, providing an important resource for Antarctic atmospheric studies.

Electromagnetic ion cyclotron (EMIC) waves play a key role in radiation belt dynamics through resonant interactions. However, their low occurrence probability, high variability, and spatial intermittency pose challenges for accurate modeling. In this study, we present a machine learning (ML)-based global EMIC wave model built on the entire data set from the Van Allen Probes mission. To capture the distinct statistical characteristics of wave occurrence and amplitude, the model is separated into two modules: an occurrence model trained using ML techniques, and a wave amplitude model sampled from observed probability distributions. The input parameters are limited to real-time or predictable variables to ensure practical applicability. Our model shows strong performance across the entire test set and demonstrates improved predictive capability over a baseline random occurrence model, particularly during quiet geomagnetic conditions. Evaluation during both quiet and active periods confirms the model's ability to represent the clustered and intermittent nature of EMIC wave activity. Furthermore, the model provides global estimates of wave power, enabling integration with radiation belt electron data and showing signatures consistent with wave-induced scattering. We found a good correlation between the global wave activity from the model and relativistic electron observation by Van Allen Probes, regardless of the availability of in situ wave observations. The modular structure of the model also allows for straightforward expansion for additional wave properties, such as wave frequency, which can be modeled independently. This flexible, event-sensitive approach offers a promising framework for data-driven radiation belt simulations and space weather applications.

Temperature variability on the synoptic scale is most directly related to human perception, and requires more attention from scientists and researchers. This study quantifies the day-to-day temperature variability (DTD) as the absolute value of the difference between the air temperatures from two consecutive days, and analyzes the spatiotemporal variations using meteorological station observations and four reanalysis data sets. Across China, the annual cycle of DTD ranges from 1.4 to 2.5°C with distinct seasonality: spring and winter exhibit a larger DTD magnitude than summer. There are large heterogeneities among different river basin regions, with the highest annual mean DTD observed in the Songliao River Basin (2.18°C) and the lowest in the Southwest River Basin (1.08°C). For the reanalysis data sets, the DTD values from JRA55 are closest to the observations, with a largest correlation of 0.98 in Southwest River Basin and smallest RMSE of 0.01°C in Yangtze River Basin. The DTD trends from JRA55 and ERA5 are comparable to those from the observational data. Further analysis of the top 1, 5 and 10 day-to-day temperature variabilities (DTDmax1, DTDmax5, DTDmax10) reveals that these large DTD values are characterized by rapid cooling, with DTDmax1 exhibiting a fluctuation range of 3.29°C–14.59°C. Additionally, DTDmax1 occurs mainly in spring and winter, with the occurrence date becoming earlier (−8.8 days/10y, p < 0.05) over the study area between 1980 and 2022. These results enhance our understanding of temperature changes at the synoptic scale, providing better services for warning and mitigating disasters.

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.