Goddard College

Hydrodynamic models in coastal and estuarine systems are typically constrained by sparse bathymetry, boundary, and validation data, especially in regions where field campaigns are costly or impractical. Here we develop and test a fully satellite-driven framework for hydrodynamic modeling in South Africa’s Langebaan Lagoon without using any local in situ measurements. Bathymetry is derived by training multispectral Sentinel-2 reflectance against ICESat-2 ATL24 photon-derived depths using an XGBoost model optimized with Bayesian search. The final satellite derived bathymetry reproduces independent ATL24 points with RMSE = 0.45 m and R 2 = 0.97. This bathymetry was used in a depth-averaged Delft3D Flexible Mesh model driven at the open boundary by TPXO tidal harmonics and by ERA5 winds. We validate modeled water surface elevation against 16 SWOT low-rate (250 m, unsmoothed) passes in 2023. SWOT–model comparisons yield an overall RMSE of 0.11 m and R 2 = 0.61, with typical point differences &amp;lt;0.10 m ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="m1"> <mml:mrow> <mml:mo>∼</mml:mo> </mml:mrow> </mml:math> 5% of the 2 m tidal range), and showed consistent spatial gradients in water level from the offshore boundary, through Saldanha Bay, and into the lagoon. At the offshore boundary, TPXO and SWOT sea surface heights agree closely ( R 2 = 0.86). A <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="m2"> <mml:mrow> <mml:mo>∼</mml:mo> </mml:mrow> </mml:math> 26 min phase lag, determined using a lag-correlation analysis, reduces the TPXO–SWOT RMSE from 0.18 m to 0.11 m, indicating that phase differences explain some of the mismatch, with remaining differences likely linked to non-tidal signals. Our results demonstrate that combining passive optical, photon-counting LiDAR, radar interferometry, and global tidal/atmospheric models enables robust, transferrable hydrodynamic modeling in data-scarce coastal systems, offering a cost-effective pathway for monitoring.

The Hyper-Angular Rainbow Polarimeter 2 (HARP2) on the NASA Plankton Aerosol Cloud ocean Ecosystem (PACE) mission is a wide field of view imaging polarimeter instrument designed for highly accurate and resolved cloud observations. HARP2 is uniquely sensitive to the polarized cloudbow, a ring-like structure in polarized light that appears above liquid water clouds. The structure of the cloudbow encodes information about the droplet size distribution, which is a critical link between cloud microphysical and radiative properties. Matching a multi-angle measurement of the cloudbow to Mie scattering predictions allows for a retrieval of important cloud properties: droplet effective radius and variance. HARP2 is the first instrument of its kind suitable for this retrieval at 5 km spatial resolution. Its wide swath facilitates global coverage of polarimetric measurements in 2 days, making it a uniquely powerful tool for studying cloud microphysics. This paper briefly presents the HARP2 instrument, demonstrates its retrieval capabilities, and discusses future science that it makes possible.

This paper introduces a novel machine learning framework for dynamic risk-based asset allocation that addresses fundamental limitations in traditional portfolio optimization methods. The proposed architecture integrates Long Short-Term Memory networks for volatility forecasting with differentiable risk budgeting layers and regime-switching mechanisms, enabling end-to-end training of portfolio weights under adaptive risk constraints. Unlike conventional approaches that rely on static risk budgets and historical covariance estimates, our methodology dynamically adjusts risk targets based on real-time market indicators, including volatility expectations, credit spreads, and yield curve dynamics. The framework achieves three primary research objectives: first, it demonstrates superior risk-adjusted performance with a Sharpe ratio of 1.38 during the out-of-sample period (2017-2022), representing a 55% improvement over traditional risk parity strategies and a 23% enhancement over contemporary deep learning approaches. Second, the architecture maintains computational efficiency through sparse attention mechanisms, scaling linearly with asset count while processing 50-asset portfolios in under 25 milliseconds. Third, the model preserves interpretability via SHAP-based risk attribution, providing transparent insights into allocation decisions across different market regimes. Empirical results reveal particularly strong performance during volatile market conditions, with maximum drawdowns reduced by 41% during stress periods compared to conventional methods. The framework's proactive risk management capabilities were evidenced during the COVID-19 crisis, where it began reducing equity exposure two weeks before the market trough, demonstrating genuine predictive ability rather than reactive adjustment. Robustness checks confirm performance persistence under varying transaction costs, rebalancing frequencies, and alternative risk measures. These findings establish a new paradigm for portfolio optimization that successfully bridges theoretical finance with practical implementation. The framework's ability to navigate complex market environments while maintaining computational efficiency and interpretability suggests readiness for widespread institutional adoption. This research contributes to the evolving literature on differentiable finance while providing portfolio managers with a robust tool for constructing adaptive, risk-aware investment strategies.

We present a deep learning approach for estimating aerosol optical depth (AOD) using normalized radiance data from GEMS, the first geostationary hyperspectral satellite dedicated to environmental monitoring. The method employs TabNet, an interpretable neural network architecture that combines hyperspectral radiance measurements with meteorological and ancillary inputs. The proposed model demonstrated strong agreement with reference data under multiple cross-validation scenarios (R2 = 0.89–0.93; within expected error: 75.1–84.6%), substantially outperforming the existing GEMS Level-2 AOD product. Our results reveal the model’s effectiveness in capturing dynamic aerosol distributions across the Asia–Pacific region, especially during high-impact events such as dust storms and biomass burning. This framework offers a promising alternative for satellite-based AOD retrieval, enhancing support for air quality assessment and atmospheric research.

Accurate forecasting of PM₂.₅ (particulate matter ≤ 2.5 μm) is essential for effective air quality management, particularly in urban areas like Delhi, which frequently experience severe pollution episodes. This study evaluates the predictive capabilit

<strong class="journal-contentHeaderColor">Abstract.</strong> The lidar backscattering properties of Asian dust particles, namely the lidar ratio (𝑆) and backscattering depolarization ratio (&delta;), were studied using a discrete dipole approximation (DDA) model. The three-dimensional morphology of the dust particles was reconstructed in fine detail using the focused ion-beam (FIB) tomography technique. An index based on the symmetry of the scattering phase matrix was developed to assess the convergence of random orientation computation using DDA. Both the 𝑆 and &delta; exhibit an asymptotic trend with dust particle size: the 𝑆 initially decreases while the &delta; increases with size, before both approach their asymptotic values. The lidar properties were found to have statistically insignificant dependence on effective sphericity. The presence of strongly absorbing minerals, such as magnetite, can greatly reduce the dust's single-scattering albedo and &delta;. Utilizing the robust asymptotic trend behavior, two parameterization schemes were developed: one to estimate the &delta; of a single dust particle given its size, and the other to estimate the &delta; of dust particles with a lognormal particle size distribution given the effective radius. The parameterization scheme was compared with results based on the TAMUdust2020 database, showing hexahedrals to reasonably represent realistic geometries with similar physical properties.

Achieving the recommended daily level of physical activity (PA) in adolescent boys and girls helps improve their cardiovascular health, maintain a healthy body weight, and promote mental health. The purpose of our study was to determine whether there were statistically significant differences between boys and girls aged 12 to 15 years with respect to age and amount of PA on different days of the week. The sample included 606 7th, 8th, and 9th grade students from four Slovenian elementary schools who participated in the study by completing a questionnaire (IPAQ; International Physical Activity Questionnaire, 2006). We used a quantitative research approach and the method was causal-non-experimental. The Shapiro-Wilk test was used to determine whether the data were normally distributed at a statistically significant level of risk (p &gt; 0.05). Because of the non-normal distribution of the variables (p &lt; 0.001), the non-parametric Mann-Whitney test was used to determine statistically significant differences between genders, and the Kruskal-Wallis test was used to determine statistically significant differences between age groups. Boys were statistically significantly more active than girls on all days of the week except Wednesday (Mon.: p=0.018; Tues.: p=0.001; Thurs.: p=0.020; Fri.: p=0.015, Sat.: p=0.012; Sun.: p=0.001). Significant differences in PA levels between seventh-, eighth-, and ninthgraders occurred only on Sundays (p&lt;0.001), with PA levels changing with student age. The study's findings that adolescent boys are statistically significantly more active than girls raise concerns about gender differences in PA participation during this critical developmental period. Such differences may contribute to long-term health consequences and perpetuate gender inequalities in overall fitness and well-being. The observed differences in PA between boys and girls during adolescence highlight the need for targeted interventions and strategies to promote and improve PA among girls. Addressing the underlying factors that contribute to lower activity levels among girls is essential to promoting a more equitable and inclusive approach to PA promotion.

Contrast adaptation is a fundamental visual process that has been extensively investigated and used to infer the selectivity of visual cortex. We recently reported an apparent disconnect between the effects of contrast adaptation on perception and functional magnetic resonance imaging BOLD response adaptation, in which adaptation between chromatic and achromatic stimuli measured psychophysically showed greater selectivity than adaptation measured using BOLD signals. Here we used magnetoencephalography (MEG) recordings of neural responses to the same chromatic and achromatic adaptation conditions to characterize the neural effects of contrast adaptation and to determine whether BOLD adaptation or MEG better reflect the measured perceptual effects. Participants viewed achromatic, L-M isolating, or S-cone isolating radial sinusoids before adaptation and after adaptation to each of the three contrast directions. We measured adaptation-related changes in the neural response to a range of stimulus contrast amplitudes using two measures of the MEG response: the overall response amplitude, and a novel time-resolved measure of the contrast response function, derived from a classification analysis combined with multidimensional scaling. Within-stimulus adaptation effects on the contrast response functions in each case showed a pattern of contrast-gain or a combination of contrast-gain and response-gain effects. Cross-stimulus adaptation conditions showed that adaptation effects were highly stimulus selective across early, ventral, and dorsal visual cortical areas, consistent with the perceptual effects.

Not every item in a stimulus ensemble equally contributes to the perceived ensemble average. Rather, items with feature values close to the ensemble mean (inlying items) contribute stronger to the estimate compared to those items whose feature values are further away from the mean (outlying items). This non-uniform weighting process, named robust averaging, has been taken as evidence that human observers are not optimally integrating sensory information. Here, we argue that robust averaging naturally emerges from an optimal integration process where the sensory en- coding is efficiently adapted to the overall ensemble statistics in the experiment. We demonstrate that such a model can accurately fit several existing datasets showing robust averaging behavior of low-level stimulus features such as orientation. Across various feature domains, our model accurately predicts subjects’ estimation accuracy and non-uniform weighting profile, and both their dependency on the specific stimulus distribution in the experiments. Our results suggest that the human visual system can form efficient sensory representations on short time-scales.