Manuel Pulido

Understanding pedestrian dynamics is critical for mitigating crowd-related risks and improving public safety. In this work, we propose a novel data-driven mesoscopic modeling framework that combines the kinetic theory of active particles (KTAPs) with data assimilation techniques, thereby offering a data-driven learning approach applicable to collective systems with learning and decision-making behavior, in line with recent KTAP developments in this area. The framework uses an ensemble Kalman filter to sequentially estimate the time-dependent spatial distribution of pedestrians and model parameters by fusing observations with the mesoscopic forward model state. Through a series of twin experiments, we show that the panic factor—a key behavioral parameter—is identifiable within this framework. We also evaluate the robustness of the approach by assimilating synthetic observations generated by an agent-based model (ABM). This setup introduces structural model error because the ABM is governed by microscopic rules that differ fundamentally from the mesoscopic kinetic equations. Despite this discrepancy, the ensemble Kalman filter, through the observation-based innovation term, successfully drives the kinetic model to track the observed pedestrian density while simultaneously recovering the panic factor. In this framework, observations act as a physical constraint on the evolution of the kinetic model. Crowd-dynamics models, including ABMs and kinetic models, often rely on phenomenological terms to describe social interactions, with parameters that are highly uncertain. Our findings indicate that in such systems, free-running simulations inevitably may diverge from the true state, whereas an online data-driven approach effectively constrains the system’s trajectory to its underlying dynamical manifold.

El núcleo agrı́cola del sudeste de la Provincia de Córdoba ocupa una región de planicies con aguas subterráneas cercanas a la superficie, capaces de complementar los aportes pluviales en periodos en los que son deficitarios. Este trabajo investiga cómo impactó la última sequı́a 2019-2023 sobre la humedad del suelo superficial, la humedad del suelo en profundidad y el nivel freático, en 8 sitios cercanos a la localidad de Marcos Juárez. Para ello, analizamos observaciones puntuales de nivel freático (NF) y cuatro productos (SMAP-L4, SMOS-L3-L4, VST-CONAE, NOAH-SMAP) que informan la humedad del suelo en superficie (HSS) y profundidad (HSP). Sin que la serie de valores absolutos de precipitación mostrara una tendencia marcada entre junio 2018 y diciembre 2023, las series de valores absolutos de NF, HSS y HSP declinaron lineal y significativamente, evidenciando retardos entre las manifestaciones meteorológicas, agrı́colas e hidrológicas de la sequı́a. En promedio, el NF descendió 0.53 m/año, pasando de -1.42 m (enero 2019) a -4.85 m (octubre 2023), con marcas <-6 m en algunos sitios, no registradas desde principios del siglo. Tres de cuatro productos detectaron que la HSS presenta mayor tendencia, rango dinámico y variabilidad temporal que la HSP, en consonancia con el conocimiento general existente. Distintos análisis sobre estos productos muestran: mayor humedad para VST-CONAE y NOAH-SMAP que para SMAP-L4 y SMOS-L3-L4; acuerdo en las anomalı́as mensuales positivas antes del 2020 y negativas después del 2021; mayor correlación temporal de las anomalı́as mensuales medias en las combinaciones SMAP-L4 con SMOS-L3-L4 y SMAP-L4 con NOAH-SMAP. El NF se asoció significativamente a HSP mientras que la precipitación solo lo hizo con HSS. Observamos correlación no lineal entre el NF y HSP en casi todos los sitios y productos, acentuada cuando la HSP precede en un mes al NF. Los rendimientos del maı́z y la soja disminuyeron en general hacia el final del periodo, cuando la sequı́a se hizo mayormente presente. Sin embargo, hubo excepciones para la soja, donde el NF, aun con una tendencia de descenso, se mantuvo cerca de la superficie, minimizando el impacto del déficit sobre el rendimiento.

An offline methodology is applied to estimate parameters of a subgrid-scale non-orographic gravity-wave scheme using observations from constant-level balloons. The approach integrates the Ensemble Kalman Filter (EnKF) with an iterative parameter estimation method based on the expectationmaximization (EM) algorithm. The meteorological fields required for the parameterization offline are taken from the ERA5 reanalysis, corresponding to the instantaneous meteorological conditions found underneath the Strateole-2 balloon observations made in the lower tropical stratosphere from November 2019 to February 2021 and October 2021 to January 2022. Compared to a direct approach that minimizes a cost function and uses Bayesian inference of parameters, our analysis demonstrates that the EnKF/EM method effectively characterizes the launching amplitudes and altitudes of the parameterized gravity waves and while quantifying their associated uncertainties. Furthermore, we illustrate how the method can help improving a scheme, specifically the results indicate that introducing a background wave activity renders the convective wave parameterization more realistic.

The extraction of information about traffic accidents from legal documents is crucial for quantifying insurance company costs. Extracting entities such as percentages of physical and/or psychological disability and the involved compensation amounts is a challenging process, even for experts, due to the subtle arguments and reasoning in the court decision. A two-step procedure is proposed: first, segmenting the document identifying the most relevant segments, and then extracting the entities. For text segmentation, two methodologies are compared: a classic method based on regular expressions and a second approach that divides the document into blocks of n-tokens, which are then vectorized using multilingual models for semantic searches (text-embedding-ada-002/MiniLM-L12-v2). Subsequently, large language models (LLaMA-2 7b, 70b, LLaMA-3 8b, and GPT-4 Turbo) are applied with prompting to the selected segments for entity extraction. For the LLaMA models, finetuning is performed using LoRA. LLaMA-2 7b, even with zero temperature, shows a significant number of hallucinations in extractions which are an important contention point for named entity extraction. This work shows that these hallucinations are substantially reduced after finetuning the model. The performance of the methodology based on segment vectorization and subsequent use of LLMs significantly surpasses the classic method which achieves an accuracy of 39.5%. Among open-source models, LLaMA-2 70B with finetuning achieves the highest accuracy 79.4%, surpassing its base version 61.7%. Notably, the base LLaMA-3 8B model already performs comparably to the finetuned LLaMA-2 70B model, achieving 76.6%, highlighting the rapid progress in model development. Meanwhile, GPT-4 Turbo achieves the highest accuracy at 86.1%.

Soil moisture is a variable of direct interest for agricultural production. However, its high spatio-temporal variability makes estimation difficult. Remote sensors aboard satellites enable for soil moisture measurements over large areas with good spatial and temporal resolution. Despite these advantages, the measurements are limited, biased, and might be spurious, requiring corrections. This work proposes the evaluation of data assimilation based on the Ensemble Kalman filter with perturbed observations to estimate soil moisture in the Argentinian Pampas under the extreme drought conditions of 2022. Two techniques for observational bias correction in soil moisture assimilation are evaluated: cumulative distribution matching and normal deviate correction. Due to the drought-induced moisture anomaly, the cumulative distribution matching corrections result in almost constant corrected observations for the entire dry season in SMAP, and in a less determinant way for SMOS. Despite this, the cumulative distribution matching technique presents the lowest root mean squared error when compared to in-situ measurements.

Many frameworks exist to infer cause and effect relations in complex nonlinear systems but a complete theory is lacking. A new framework is presented that is fully nonlinear, provides a complete information theoretic disentanglement of causal processes, allows for nonlinear interactions between causes, identifies the causal strength of missing or unknown processes, and can analyze systems that cannot be represented on Directed Acyclic Graphs. The basic building blocks are information theoretic measures such as (conditional) mutual information and a new concept called certainty that monotonically increases with the information available about the target process. The framework is presented in detail and compared with other existing frameworks, and the treatment of confounders is discussed. While there are systems with structures that the framework cannot disentangle, it is argued that any causal framework that is based on integrated quantities will miss out potentially important information of the underlying probability density functions. The framework is tested on several highly simplified stochastic processes to demonstrate how blocking and gateways are handled, and on the chaotic Lorentz 1963 system. We show that the framework provides information on the local dynamics, but also reveals information on the larger scale structure of the underlying attractor. Furthermore, by applying it to real observations related to the El-Nino-Southern-Oscillation system we demonstrate its power and advantage over other methodologies.

The impact of gravity wave drag on the Antarctic sudden stratospheric warming (SSW) in 2002 is examined through a mechanistic middle atmosphere model combined with a variational data assimilation system. Significant differences in the SSW representation are found between a model integration that uses reference gravity wave parameters and one that uses parameters estimated using data assimilation. Upon identical wave forcings at 100 hPa, the vortex breakdown may arise as either a vortex splitting event or a displacement vortex event depending on gravity wave parameters. A local enhancement of Rossby waves is found in the integration with estimated parameters, leading to a split SSW. The changes in the vortex breakdown are associated with changes in the vortex geometry caused entirely by modifying the gravity wave parameters. Gravity wave drag proved to play an instrumental role in preconditioning the stratosphere near a resonant excitation point that triggers the split SSW.

For modelling geophysical systems, large-scale processes are described through a set of coarse-grained dynamical equations while small-scale processes are represented via parameterizations. This work proposes a method for identifying the best possible stochastic parameterization from noisy data. State-of-the-art sequential estimation methods such as Kalman and particle filters do not achieve this goal successfully because both suffer from the collapse of the posterior distribution of the parameters. To overcome this intrinsic limitation, we propose two statistical learning methods. They are based on the combination of the ensemble Kalman filter (EnKF) with either the expectation–maximization (EM) or the Newton–Raphson (NR) used to maximize a likelihood associated to the parameters to be estimated. The EM and NR are applied primarily in the statistics and machine learning communities and are brought here in the context of data assimilation for the geosciences. The methods are derived using a Bayesian approach for a hidden Markov model and they are applied to infer deterministic and stochastic physical parameters from noisy observations in coarse-grained dynamical models. Numerical experiments are conducted using the Lorenz-96 dynamical system with one and two scales as a proof of concept. The imperfect coarse-grained model is modelled through a one-scale Lorenz-96 system in which a stochastic parameterization is incorporated to represent the small-scale dynamics. The algorithms are able to identify the optimal stochastic parameterization with good accuracy under moderate observational noise. The proposed EnKF-EM and EnKF-NR are promising efficient statistical learning methods for developing stochastic parameterizations in high-dimensional geophysical models.