Abstract: Background In precision agriculture, especially in hydroponic systems where environmental management affects both production and resource efficiency, interpretable modeling of crop phenological stages is crucial. Typical visualization-based approaches are generally expensive and ecologically dependent. To classify and visualize lettuce growth stages (Seedling, Vegetative Growth, Heading, Bolting and Senescence) using temperature, humidity, total dissolved solids (TDS) and pH, this study proposed XGB-StageVision, a sensor-powered machine learning setup combining supervised and unsupervised learning techniques. Using XG Boost, LightGBM, CatBoost, ExtraTrees, Random Forest and LSTM under an 80/20 train-test split, a dataset of 3,169 unique observations from 70 hydroponically grown lettuce plants was evaluated. Methods and Results XGB-StageVision achieved 99.86% test accuracy, a ROC AUC of 1.00, an MCC of 0.9982 and perfect precision, recall, F1 scores and specificity across all classes, performing competitively with the best baseline models, including LightGBM, which achieved statistically comparable cross validation performance. The unsupervised t-SNE visualization showed clearly distinct structural clusters across growth stages, while SHAP revealed temperature as the strongest predictor (≈75% contribution), followed by TDS. Novel Contributions SHAP, LIME and t-SNE provided global, local and structural interpretability and transparent perspectives on environmental influences regarding plant development. Statistical analysis and 10-fold cross-validation were used to further test the model’s robustness, reliability and generalizability. Conclusion XGB-StageVision provides an IoT-ready, interpretable and imaging-free system for monitoring lettuce growth stages in real time. The framework combines supervised predictive modeling with unsupervised structural visualization to ensure transparency, sustainability and high performance, supporting resource-efficient, data-driven strategic planning in contemporary hydroponic agriculture to lower monitoring costs, optimize inputs and improve yield consistency.

Abstract: In this paper, we characterize the two-weight core-EP inverse using the weighted core decomposition and establish its novel representations through the weighted g-Drazin inverse. Furthermore, we define and investigate the weighted core-EP orders within this general setting. These results subsume and extend the properties of the core-EP inverse to a broader algebraic setting.

Abstract: This study analyzes daily bank stock returns and their dynamic connectedness for nine American banks between March 18, 2022, and March 15, 2023. It examines the interrelationships between bank stock returns and three economy-wide variables: market sentiment, monetary conditions represented by the Secured Overnight Financing Rate (SOFR), and economic policy uncertainty (EPU). The analysis employs a framework that combines correlation, cointegration, and connectedness (3C), using a time-varying parameter vector autoregression (TVP-VAR) model to examine evolving forecast-error spillovers. Using a 12-variable TVP-VAR model with generalized forecast error variance decomposition, the full-system total connectedness index is 75.4%. SOFR emerges as the primary average net transmitter (+17.6 pp), reflecting funding-cost repricing amid aggressive policy tightening, whereas EPU (-28.1 pp) and VIX (-17.4 pp) are net receivers. The findings provide descriptive evidence of time-varying stress transmission across regional lenders without establishing causal predictive links to bank insolvency.

Abstract: Background/Objectives: A model explanation can still look clinically credible after the model has stopped travelling well. Hospital transfer studies rarely test that mismatch directly. We examined whether predictions and feature attributions changed together when sepsis risk models moved between two hospital systems. Methods: We analyzed the public PhysioNet/Computing in Cardiology Challenge 2019 data. The cohort included adult stays with at least 24 hourly records and no positive sepsis label during the first 12 h. We derived 115 predictors from that 12-h window, then trained elastic-net logistic regression, random forest, and XGBoost models separately in systems A and B. Each fitted model was evaluated in its source system and, without target-site refitting, in the other system. We assessed calibration, a frozen source threshold chosen for 80% sensitivity, decision curves, subgroup performance, and 1000-resample bootstrap intervals. The XGBoost audit compared global SHAP rankings, top-10 overlap, attribution direction, missingness-attribution mass, an ablation without observation indicators, and classifiers trained to identify the hospital system. Results: The cohort comprised 30,646 stays with 1843 subsequent sepsis events. XGBoost AUROC fell from 0.725 internally to 0.608 after A→B transfer and from 0.702 to 0.636 after B→A transfer. External calibration slopes were 0.431 and 0.700. SHAP-rank correlations remained high at 0.949 and 0.917, yet the top-10 Jaccard overlap was only 0.667 in either direction; the within-site bootstrap reference was 1.000. Observation indicators contributed 24.7% of external attribution mass for A→B and 7.5% for B→A. On their own, those indicators identified the hospital system with AUROC 0.998, although removing them changed external AUROC by less than 0.01. Conclusions: The two hospitals left a clear signature in what clinicians measured and what they left unmeasured. That signature altered the explanation profile even when its incremental contribution to discrimination was small. External validation should test prediction, calibration, threshold behavior, and explanation stability together.

Abstract: In this presentation, we investigate the theoretical and practical aspects of fixed point theory on metric spaces equipped with directed graphs. Aiming to relax global contraction conditions, graph related fixed point theorems provide a powerful setting for solving non-linear differential and integral equations. We establish a fixed point theorem for graph preserving self-mappings on such spaces. In addition to presenting a non-trivial example validating our main results, we demonstrate two concrete applications: the existence and uniqueness of solutions for non-linear Caputo fractional differential equations, and the existence of unique periodic trajectories in non-autonomous dynamical systems with time delay.

Abstract: Scientific laws become known under determinate conditions, yet their claim to objectivity reaches beyond any single realization. This paper develops a categorical account of that relation through meta-laws M, first-order laws L, and objective observational conditions O. A four-quadrant ontology distinguishes the relations retained by different accounts of law. Each quadrant may possess symmetry, normativity, causal organization, and statistical structure within its own domain. The question is whether it can preserve these relations across the complete family of laws and conditions. Replacing a factor by the terminal category gives the corresponding specialized quadrant; when that erasure loses an essential explanatory dependence, we interpret the specialization as ontological degeneration. The formal comparison distinguishes fixing a condition from erasing an axis: a reconstruction essentially dependent on both axes cannot factor, even up to natural isomorphism, through either axis alone, although a suitable restricted domain can compute the same universal object. Quadrant IV-C retains the complete law--condition family and its coherent global comparison.The local reconstruction F(L,O)=Ψ(R(L,O)) is related to the global meta-law by specified universal cones or cocones. Presheaf extension yields an adjunction; representability characterizes its descent to the original categories, and a finite logical model realizes the resulting joint inference. The complete functor determines a group of compatible invertible comparisons, whose action on the global reconstruction is induced by universality. Its inherited symmetry group is the quotient by the kernel of that action and agrees with the comparison group when the action is faithful. A finite calculation shows that identical law-only restrictions can yield different inherited groups. Global symmetry thus follows from the complete relational construction without an independent postulate of its action.Symmetry is the first of eight structural developments. Normativity expresses the compatibility requirements of a specified covariance structure; in a smooth bundle realization, these requirements determine the transformation law of a connection. Causality concerns the composition of interpreted temporal or intervention relations. Statistical covariance concerns coherent transport of probability laws. Historicity concerns comparisons among laws and conditions through time, including their changing domains of applicability. Modality evaluates necessity, possibility, and counterfactual claims over admitted realizations. Lewis counterfactuals receive a sphere semantics that remains meaningful without closest antecedent worlds, while the comparison with Kripke makes accessibility and valuation explicit. The distinction between composition and invertibility explains their different modal consequences.The seventh development concerns higher structure: a diagrammatic 2-functor compares transformations themselves, and a specified smooth realization expresses their coherence through path and surface transport. The eighth concerns topos structure. Presheaf and sheaf realizations give laws and their applicability an internal logical expression; a characteristic morphism represents a chosen applicability domain, and naturality governs its behavior under restriction. Scientific modeling and evidence determine which domain is appropriate. These constructions articulate the conditionality, boundedness, hierarchy, and structural relations of laws within the same account of intelligibility. Sheaf gluing connects compatible local descriptions with global realization, while solution torsors locate its obstruction. A finite circle calculation distinguishes an incompatible choice of local data from a nonzero obstruction to any compatible choice.The empirical discussion distinguishes covariance tests, obstruction tests for identified reconstruction problems, and level-evolution tests of theoretical hierarchy. Exact reconstruction, approximate recovery, and discrimination between predictions impose different requirements; comparisons with observations need common readouts and explicit error bounds. Six scientific cases examine how these distinctions bear on changes of physical theory, measurement, symmetry breaking, social reflexivity, market conditions, and effective descriptions across scales. They illustrate the framework's scientific interpretation without claiming new physical predictions from its abstract form.The philosophical argument locates objectivity in coherent relations among manifestations, with observational conditions participating in those relations. On this interpretation, scientific inquiry has always inhabited quadrant IV-C. The aspiration to a description independent of every observational condition can conceal this relation by treating O as a transparent window and a local realization as the whole. The resulting transcendental forgetting concerns conditions already at work in knowledge. Through the geometry of observational space and its dialogue with the philosophical traditions, the M–L–O account brings them into view. Its mathematical constructions establish what the specified relations preserve and what their erasure loses; its philosophical thesis is that recognizing these worldly conditions belongs to understanding objectivity itself.

Abstract: Longitudinal autism intervention review requires specialists to combine observations from caregivers, teachers, and intervention staff across settings and time. We developed a governance-aware human-in-the-loop application that uses AI to organize evidence while keeping intervention decisions with professionals. The system preserves original observations, proposes structured records with Qwen 2.5 7B, requires professional verification or correction before analytics, checks evidence adequacy, supports abstention, and produces traceable reports. Evaluation used archival ASDBank data from 11 children, 85 sessions, and 9,018 behavioral episodes. Two experienced autism specialists created the reference standard, followed by a final-codebook compliance audit. Qwen 2.5 7B achieved 41.3% merged categorical agreement on the 330-segment main batch and remained the strongest evaluated generator. The offline MARS selector returned records for 80.3% of segments but showed no clear field-level advantage over the strongest reviewable generator. A historical generative-reporting experiment failed its full governance contract (0/11), motivating deterministic operational reporting. Final controlled end-to-end testing passed all predefined governance scenarios (77/77 API tests and 9/9 UI scenarios). These findings support a practical architecture in which AI assists evidence organization, while professional review, abstention, provenance, and deterministic controls protect the longitudinal workflow.

Abstract: We use Random Matrix Theory to analyze the impact of COVID-19 on the correlation structure of the agricultural futures market, using daily returns for 56 futures from various countries around the world before, during, and after COVID. For each of the three time periods, we build correlation matrices of the daily returns, and we examine the matrices’ entries, eigenvalues, and eigenvectors. We find that agricultural futures exhibited stronger collective behavior during COVID than before or after it. We also find clustering based on geographic location, with there being one cluster in the West (North America, South America, Europe) and one in the East (Asia, Oceania), which show predominantly opposite-signed loadings relative to the dominant market mode; since pre-pandemic times, the East has gained weight in the market while the West has lost it. On top of regional clustering, we also find independent exchange-level clustering, where futures tend to be more correlated with futures in its same exchange than any other. Our results demonstrate how Random Matrix Theory can reveal both transient changes in collective behavior and longer-term structural patterns in an interconnected global market.

Abstract: For a fixed integer \( n\ge2 \), let \( \Delta_{n-1} \) denote the standard probability simplex in $\( \mathbb R^n \)$. We introduce and analyze the self-power escort transformation \( T_n:\Delta_{n-1}\to\Delta_{n-1} \) defined by \( \begin{equation*}T_n(p)_i=\frac{p_i^{p_i}}{\sum_{j=1}^n p_j^{p_j}},\qquad i=1,\ldots,n,\end{equation*} \) with the continuous boundary convention \( 0^0=1 \). Unlike the usual power escort transformation, whose exponent is an external parameter, the exponent here is the coordinate itself; equivalently, each coordinate is reweighted by \( \exp(p_i\log p_i) \). The paper proves that this elementary self-feedback rule has a rigid global dynamics. Every boundary point is immediately activated, every orbit enters an explicitly described compact core after one step, the uniform distribution is the unique fixed point, and the Hilbert log-diameter contracts exponentially after core entrance. Consequently every orbit converges exponentially to the uniform distribution, and no nontrivial periodic orbit exists. We then compute the complete linear spectrum and second-order normal form at the uniform state, obtaining a sign transition between the binary and higher-dimensional systems. For \( n\ge4 \), the post-core Hilbert contraction constant is sharpened to a two-variable min--max problem involving only the smallest and largest coordinates. We also classify the entire one-step uniformization fiber \( T_n^{-1}(\nu_n) \): besides the vertices and the uniform point, each dimension \( n\ge3 \) has exactly one nonuniform interior orbit type, up to permutation. Finally, in the binary case we prove global contraction in logit coordinates and an arithmetic escape phenomenon: algebraic irrational inputs become transcendental after one step, while rational nonuniform inputs become transcendental after two steps.

Abstract: We disprove the conjecture of Camacho, Khudoyberdiyev and Omirov that \(x_i^2 = \sum_j b_{ij}x_j\) has a solution with every coordinate nonzero for every nonsingular complex matrix \(B\). Dimension three is the smallest in which this conjecture fails. In that dimension, all counterexamples are equivalent under simultaneous coordinate permutations and nonzero diagonal rescalings. The exceptional locus in \(\mathrm{GL}_3(\mathbb{C})\) is the disjoint union of six smooth, closed, three-dimensional algebraic tori, each a complete intersection of codimension six. Every exceptional matrix has exactly seven nonzero entries, the minimum possible for a counterexample. A regular three-dimensional evolution algebra with at most six nonzero structure constants has a full-support idempotent, and this bound is sharp. We also give polynomial equations characterizing the exceptional locus. The proof uses elimination, traces in an Artinian complete intersection, and local multiplicities.

Abstract: Childhood sexual abuse (CSA) is reliably associated with adult sexual dysfunction, yet the field has historically framed survivors' sexuality in terms of damage rather than recovery. This narrative review asks whether sexual health recovery can occur after childhood abuse, and what trauma-informed care contributes to it. Evidence from meta-analyses, psychophysiological studies, longitudinal cohorts, and randomized treatment trials is synthesized around three questions: the strength of the CSA–dysfunction association, the mechanisms linking them, and the responsiveness of sexual function to intervention. Meta-analytic evidence confirms the association (odds ratio 1.68, 95% CI [1.49, 1.87]), stronger among women than men, although findings are heterogeneous and moderated by comorbidity, abuse characteristics, and family context. Five mechanisms recur: posttraumatic stress with blunted genital arousal and altered cortisol response; dissociation during sexual activity; cognitive-affective processes centered on distraction and avoidance; negative sexual self-schemas and low sexual self-esteem; and relational difficulties including insecure attachment. A reframing of dysfunction as protective — avoiding re-traumatization, regulating closeness, restoring a sense of control — implies that these adaptations must be understood before function is restored. Recovery is nonetheless well documented: roughly one third of sexually abused children show no symptoms and two thirds recover within 12–18 months; latent class analysis of maltreated youth identifies resilient, survival, and improving sexual health trajectories; and mindfulness-based sex therapy and schema-focused expressive writing improve sexual function, with abused women sometimes improving more than non-abused women. Recovery is therefore not automatic but facilitated, and trauma-informed care is the condition under which it occurs.

Abstract: Natural selection is credited for giving direction to evolution. But it is also described as a filter acting on variation supplied by mutation, development and chance. These views highlight different regimes, not rival truths. In this paper, I ask when selection actually determines evolutionary direction. I define determination counterfactually, i.e. an outcome is selection determined if it recurs across replicate histories under the same selective regime and becomes much less likely when selection factor is removed. Here, I argue that four conditions must hold jointly- (i) selection must outweigh drift, (ii) the selective regime must persist, (iii) variation should not be biased enough to overturn selection's ranking and (iv) landscape must allow the favored route to be travelled. Each failure hands direction to a distinct process. Determination also falls as outcomes are described more firmly, from function to phenotype to sequence. Finally, six testable predictions are stated.

Abstract: Background/Objectives: Postural control undergoes significant maturation from adolescence to adulthood. While is promising as a means of enhancing postural stability in adults, its effect in adolescents, whose postural control is poorer, remains to be determined. This study aimed to assess the effects of a yoga intervention on postural balance in healthy adolescents and adults. Methods: Eighteen adolescents and 18 young adults completed eight two-hour Hatha yoga sessions (one per week). Postural control was assessed using a stabilometric platform to measure center of pressure (CoP) parameters. These parameters were assessed in eyes-open (EO) and eyes-closed (EC) conditions. Results: Adolescents showed higher baseline oscillations than adults. While the yoga intervention demonstrated a significant main effect of Time in the EO condition for all CoP parameters and in both groups, no difference appeared in the EC condition. A significant Group × Time interaction for CoP velocity in EO showed significantly decreased CoP velocity in adolescents but not in adults. Conclusions: An eight-week yoga program significantly improved postural control in healthy adolescents and adults in the EO condition. Adolescents exhibited CoP velocity decrease after yoga intervention, whereas adults did not. These findings advocate for integrating yoga into physical education so as to support postural development, especially during adolescence.

Abstract: Most existing encrypted-traffic classification methods treat each traffic flow as an independent sample, overlooking the relationships shared by similar flows. This can amplify within-class variation and increase overlap between classes, particularly when their distinguishing characteristics are subtle and labeled data are limited. To address this problem, we propose PGCL-CG, a prototype-guided semi-supervised graph contrastive learning framework. We represent flow-level statistical features as a sparse, weighted k-nearest-neighbor graph and use GraphSAGE to aggregate information from neighboring flows and to capture their underlying structural relationships. The encoder is pretrained using self-supervised graph contrastive learning to obtain stable representations. We use a student-teacher architecture to produce more stable pseudo-labels and apply hierarchical density-based spatial clustering to filter out low-confidence assignments. We conduct extensive experiments on the two open datasets, demonstrating that PGCL-CG outperforms competitive baselines and remains robust even when only a small fraction of labels (e.g., 20% or 40%) is available.

Abstract: Ring chromosomes (RCs) are rare structural chromosomal abnormalities caused by terminal breaks and fusion of chromosome ends, occurring in approximately 1 in 50,000 live births. They are often associated with “ring chromosome syndrome,” characterized by growth restriction, mild dysmorphism, and variable intellectual disability, even without large deletions. Ring chromosome 2 is exceptionally rare, with fewer than 20 reported cases and limited molecular characterization. We report a 4-year-old male with proportional growth restriction, mild facial dysmorphism, and delayed expressive speech. Cytogenetic analysis identified a mosaic ring chromosome 2, and fluorescence in situ hybridization demonstrated preserved subtelomeric regions. However, array comparative genomic hybridization revealed a 46 kb deletion at 2p25.3 involving the protein-coding gene FAM110C. This case broadens the phenotypic and molecular spectrum of ring chromosome 2 and underscores the value of combining cytogenetic and high-resolution genomic techniques to detect cryptic imbalances contributing to clinical presentation.

Abstract:

Adsorption isotherm parameters are widely used to characterize biosorbent performance and are often interpreted as intrinsic properties of the adsorbent. However, the extent to which these parameters depend on biosorbent concentration remains insufficiently understood. Here, the adsorption of Ni²⁺ and Cd²⁺ by Chlorella vulgaris was systematically investigated over a broad range of biomass concentrations by determining complete adsorption isotherms independently at each biomass level. Increasing biomass concentration consistently decreased the specific adsorption capacity and shifted the adsorption isotherms toward lower apparent saturation capacities for both metal ions. Langmuir and Freundlich models were fitted using both linearized and nonlinear regression. Despite differences in the absolute parameter values obtained by the fitting approaches, all models revealed the same biomass-dependent trend. In particular, the Langmuir maximum adsorption capacity (q_max) and Freundlich adsorption coefficient (K_F) decreased systematically with increasing biomass concentration. These results demonstrate that biomass concentration influences the experimentally observed adsorption behaviour and the isotherm parameters derived from it. Consequently, q_max and K_F should not be interpreted as universal material constants without considering the biomass concentration at which they were determined. Biomass concentration should therefore be reported and considered when comparing biosorption isotherms across independent studies.

Abstract: Non-motor symptoms drive much of the disability and quality-of-life loss in Parkinson's disease (PD), yet they remain poorly served by dopaminergic therapy. Omega-3 polyunsaturated fatty acids (n-3 PUFAs), principally eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), are biologically plausible modifiers of these symptoms through effects on membrane composition, neurotransmission, and the resolution of inflammation, but the relationship between measured blood n-3 status and the PD symptom profile has not been systematically mapped. Methods: In a retrospective cross-sectional clinic cohort, patients with PD who completed a baseline Patient-Reported Outcomes in Parkinson's Disease (PRO-PD) assessment and a Cleveland HeartLab Parkinson's Disease Panel between 2019 and 2024 were analyzed (n = 235). Analyses proceeded in three tiers: (i) an agnostic correlation screen of the full blood panel against total PRO-PD; (ii) symptom-resolved linear models of each fatty-acid marker against each of the 35 PRO-PD items, adjusted for age, gender, and years since diagnosis; and (iii) categorical below/normal/above models using laboratory reference windows. Results: Omega-3 Total and OmegaCheck were among the strongest negative correlates of total PRO-PD across the entire blood panel (ρ = −0.19, p = 0.01), while the arachidonic acid (AA)/EPA and omega-6: omega-3 ratios were among the strongest positive correlates (ρ = 0.22 and 0.19). Symptom-resolved mapping localized this signal: the largest and most consistent associations fell in the affective (anxiety, depression), motivational (loss of interest, motivation), and cognitive (memory, comprehension) domains, with comparatively sparse associations across cardinal motor items. DHA showed the largest effect estimates (e.g., depression β = −4.82, p = 0.0005; loss of interest β = −5.37, p = 0.0003). Conclusions: In this clinic cohort, low blood n-3 status was associated with greater symptom burden, and that burden was concentrated in neuropsychiatric rather than motor domains. Findings are cross-sectional and do not establish that supplementation improves symptoms; they do identify the endpoints on which interventional trials are most likely to detect an effect.

Abstract: Clozapine remains the only antipsychotic with proven superiority for treatment-resistant schizophrenia, yet its use is constrained by marked interindividual variability in both response and toxicity and by a burdensome adverse-effect profile spanning agranulocytosis, myocarditis, and metabolic syndrome. Three decades of pharmacogenetic research have sought genetic markers capable of predicting, before treatment begins, who will respond and who will suffer harm. This narrative review critically re-examines that record across three domains: pharmacokinetics (metabolism and transport), pharmacodynamics (treatment response), and adverse drug reactions, before evaluating the gap between research findings and clinical implementation. Pharmacokinetic pharmacogenetics has been the field’s most consistent disappointment: CYP1A2 genotype contributes little beyond smoking status, and CYP2C19 and ABCB1 variants exert only modest effects on plasma concentrations. Pharmacodynamic pharmacogenetics has fared no better; candidate genes in the serotonergic and dopaminergic systems, including the celebrated six-single-nucleotide-polymorphism response model, have failed replication in genome-wide and independent cohorts. Only the adverse-reaction domain has yielded durable findings: HLA-DQB1 126Q, HLA-B 158T, and population-specific alleles such as HLA-B*59:01 are biologically coherent and replicated correlates of clozapine-induced agranulocytosis, yet their predictive values remain too low to alter mandatory hematological monitoring. This review identifies five recurring methodological failures underlying this pattern of underdelivery: chronically inadequate sample sizes, phenotype heterogeneity, inadequately handled population stratification, conflation of biological plausibility with clinical evidence, and a translational logic gap between statistical association and actionable prediction. No genetic marker or panel currently meets the evidentiary threshold required to guide clozapine initiation, dosing, or monitoring. Therapeutic drug monitoring informed by smoking status already delivers much of what pharmacogenetics promises, at lower cost and with more direct capture of the dominant determinants of clozapine exposure. Large, adequately powered, trans-ancestral genome-wide studies, evaluated against prospective clinically meaningful outcomes, are needed before genetic testing can be considered ready for clozapine prescribing decisions.

Abstract: The increasing use of lithium-ion batteries (LIBs) has increased the need for physical pre-treatment methods capable of recovering electrode materials prior to downstream recycling processes. Mechanical separation can provide an effective initial removal of electrode coatings; however, residual active material may remain attached to the metallic current collectors, limiting the recovery of clean electrode fractions. In this work, a sequential mechanical–ultrasonic approach is investigated for the recovery of electrode materials from spent commercial 18650 lithium-ion battery cells. The proposed approach combines an initial mechanical removal step with short ultrasonic treatment cycles applied to the residual electrode coating. Mechanical activation between ultrasonic cycles was used to promote the detachment of remaining material while limiting prolonged exposure of the metallic current collectors to ultrasonic treatment. Preliminary continuous ultrasonic experiments showed that extended treatment increased degradation of the current collectors without producing a proportional improvement in coating removal, providing the basis for the sequential procedure. The recovered fractions were characterized to evaluate their composition, structural characteristics, and degree of separation from the metallic current collectors. A mass balance accounted for 14.58 g of recovered cathode material, while a 0.28 g difference remained unaccounted for and could not be experimentally assigned to a specific loss mechanism. X-ray diffraction confirmed chemically distinguishable cathode fractions, while energy-dispersive spectroscopy of the ultrasonic-recovered fraction showed a low average aluminum content of 0.21 wt.%. The results demonstrate that short, mechanically assisted ultrasonic treatments can effectively remove residual electrode material while limiting collector deterioration. These findings support the use of ultrasound as an assisting step following mechanical separation, rather than as a standalone delamination process, for the physical recovery of electrode materials from spent lithium-ion batteries.

Abstract: Retrospective mapping of aboveground biomass (AGB) using long satellite archives offers a valuable approach for assessing post-fire biomass change, but contemporary calibration does not necessarily ensure reliable transferability to historical conditions. The Global Ecosystem Dynamics Investigation (GEDI), a spaceborne lidar mission that provides observations of forest structure and biomass estimates, offers a potential source of contemporary reference data for model calibration. We evaluated this issue using a GEDI-calibrated Landsat workflow for the landscape affected by the 2009 Churchill wildfire in southeastern Australia. Random forest models combined optical, environmental and pre-fire predictors using 4,746 GEDI-linked reference observations from 2020–2023. Evaluation included reference-provenance auditing, spatial cross-validation across nine blocks with 300 m training exclusions, forward temporal and spatiotemporal tests within 2020–2023, paired predictor ablation, conditional-error diagnostics and an external field comparison. All reported provenance checks passed, supporting reference-processing integrity. The full model outperformed the training-mean baseline and reduced configurations in every block, although improvements were modest. Predictions showed a compressed range, with a prediction-to-reference standard-deviation ratio of 0.362 and regression slope of 0.104, overpredicting low reference biomass and underpredicting high reference biomass. Temporal and spatiotemporal transfer from 2020 to 2023 failed to outperform their corresponding training-mean baselines. Annual predictions declined after the fire and subsequently increased, but biomass loss and recovery were not independently validated. These findings support using the maps as descriptive reconstructions and highlight the need for independent structural observations across disturbance stages before quantitative recovery assessment or carbon accounting.