KMKevin Monsálvez-Pozo

Kevin Monsálvez-Pozo

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Kevin Monsálvez-Pozo has a total of 40 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Evaluation of ERA5, ERA5-Land, CERRA and NEWA datasets in reproducing observed near-surface wind speeds across Spain

    by Nuria P. Plaza-Martin, Shalenys Bedoya-Valestt, Marcos Martinez-Roig, Kevin Monsalvez-Pozo, ÀNgela Manrique-Alba, Francisco Granell-Haro, Miguel Andres-Martin, Cesar Azorin-Molina, Juli G. Pausas, Alejandro Royo-Aranda, et al.


  • Open Access

    AFNO-based downscaling of global air pollution fields

    by Kevin Monsalvez-Pozo, Francisco Granell-Haro, Marcos Martinez-Roig, Víctor Galván Fraile, Nuria P. Plaza-Martín, Martin Otto Paul Ramacher, Johannes Bieser, Johannes Flemming, Miha Razinger, Paula Harder, et al.

    Air pollution, particularly fine particulate matter (PM2.5), poses a significant risk to public health, necessitating accurate high-resolution monitoring. While global Chemical Transport Models (CTMs) like the Copernicus Atmosphere Monitoring Service (CAMS) provide continuous worldwide coverage, their coarse spatial resolution (~40 km) limits their utility for assessing local exposure relative to regional models (~10 km) that are restricted to specific domains, such as Europe. To bridge this gap, we present a novel deep learning approach for global downscaling of pollutant concentrations based on the Adaptive Fourier Neural Operator (AFNO), benchmarking its performance against a standard U-Net baseline.We adapted the Modulated AFNO architecture for spatial super-resolution, using low-resolution CAMS Global PM2.5 and dynamic meteorological fields (wind, temperature, dew point, boundary layer height). A key innovation is integrating these inputs with high-resolution static data: orography and population density. We demonstrate that directly inputting static features into the network backbone outperforms separate spatial conditioning, effectively leveraging the Fast Fourier Transform to capture long-range dependencies while respecting local physical constraints.The model was developed using daily forecasts from 2020 to mid-2025. Training used a sequential split into 2021–2024, preserving 2020 (COVID-19 anomalies) and 2025 as a held-out test set. The model effectively reconstructed fine-scale details and corrected global model biases. Verification against European Environment Agency observations (2020) confirmed performance comparable to high-resolution CAMS Europe regional forecasts. Crucially, the AFNO model consistently outperformed the U-Net baseline and traditional linear interpolation in spatial correlations and error rates. Finally, transferability tests in North America (AirNow data) confirmed the model generalizes effectively to unseen regions, maintaining lower errors than both the original global forecast and the baseline.

  • Open Access

    Interpretable Swin Transformer–Based Downscaling of PM2.5 Air Pollution Field

    by Marcos Martínez-Roig, Francisco Granell-Haro, Kevin Monsalvez-Pozo, Nuria P. Plaza-Martin, Victor Galván Fraile, Paul Ramacher Martin Otto, Johannes Bieser, Johannes Flemming, Paula Harder, Miha Razinger, et al.

    Fine particulate matter (PM2.5) is one of the most harmful air pollutants, posing severe risks to human health and contributing significantly to premature mortality worldwide. Accurate high-resolution monitoring and forecasting of PM2.5 are therefore essential for air quality management, public health assessment, and the design of effective mitigation policies. However, operational atmospheric composition models such as the Copernicus Atmosphere Monitoring Service (CAMS) provide global fields at relatively coarse spatial resolution (~40 km), limiting their ability to represent local-scale pollution patterns driven by complex interactions between emissions, meteorology, and topography. Higher-resolution regional CAMS products (~10 km) partly address this limitation but are computationally expensive and are restricted to specific geographical domains, mainly Europe. As a result, high-resolution information remains difficult to obtain consistently at the global scale.In this work, we present a deep learning–based super-resolution approach to downscale PM2.5 concentration fields from 40 km to 10 km resolution, bridging the gap between global model outputs and regional-scale applications. The proposed approach is based on a SwinFIR architecture, a hierarchical Vision Transformer that leverages shifted window self-attention to efficiently capture multiscale spatial dependencies. The model ingests multiple low-resolution dynamic variables from CAMS, including PM2.5, 2-meter temperature (T2M), 10-meter wind speed components (U10, V10), dewpoint (D2M) and boundary layer height (BLH), providing both chemical and meteorological context. In addition, high-resolution static data, such as orography and population, are introduced through a secondary branch, enabling the model to condition the super-resolutionprocess on fine-scale geographical features that strongly influence pollutant distributions. The output consists of high-resolution (10 km) PM2.5 fields. Model performance is evaluated using both a held-out test period and independent ground-based PM2.5 observations from the European Environment Agency.Results show that the model effectively reconstructs fine-scale PM2.5 structures and reduces biases present in the global forecasts. Verification against ground-based observations indicates that the model achieves performance comparable to high-resolution CAMS Europe regional forecasts. The proposed SwinFIR model consistently outperforms a carefully optimized state-of-the-art U-Net baseline across multiple evaluation criteria, including error metrics, spatial correlation, and structural consistency. These improvements reflect the ability of self-attention mechanisms to capture long-range spatial interactions that are difficult to model using purely convolutional approaches.Beyond predictive performance, we also focus on interpretability. Feature importance analyses quantify the relative contribution of each input variable, demonstrating that static inputs used play a key role in the downscaling process. Attention maps further reveal that the model focuses on physically meaningful events, including high-concentration peaks and regions of strong wind, indicating physically consistent behavior.Finally, transferability was assessed by applying the model to North America, a region unseen during training. Evaluation against AirNow observations shows reasonable generalization performance, while highlighting the need for further research to improve robustness and extrapolation to unseen regions.Overall, this study demonstrates the potential of Transformer-based architectures for data-driven downscaling of atmospheric composition fields, providing both improved accuracy and enhanced physical interpretability. The proposed framework offers a promising tool for high-resolution airquality applications based on global model outputs.

  • Open Access

    TACO: Operationalizing AI-Ready EO datasets

    by Oscar J. Pellicer-Valero, Cesar Aybar, Mikolaj Czerkawski, Carmen Oliver, Kevin Monsálvez, Julio Contreras, Gustau Camps-Valls

    The field of Artificial Intelligence for Earth Observation (AI4EO) currently suffers from significant data friction, especially when moving Petabyte-scale archives from cloud object storage to High Performance Computing (HPC) nodes. We present the TACO (Transparent Access to Cloud-Optimized datasets), a production-grade standard designed to replace file-centric legacy workflows with a high-throughput streaming paradigm.We showcase the practical implications of this architecture for the deployment of geospatial Foundation Models (FMs), by running pretrained FMs on downstream inference tasks (such as semantic segmentation or land-cover classification) directly on arbitrary samples of arbitrary cloud-hosted datasets, quickly, and without the need for local staging or any specific preprocessing. TACO bridges the gap between static cloud archives and dynamic HPC processing, allowing seamless, scalable AI4EO workflows, and fulfilling the so far unfulfilled promise of FMs of "train once, apply everywhere".References: Cesar Aybar, et al. (2025). The Missing Piece: Standardising for AI-ready Earth Observation Datasets. Poster at TerraBytes-ICML 2025 Workshop. Vancouver, CanadaTACO Foundation. (2025, November 21). The TACO specification (Version 2.0.0). https://tacofoundation.github.io/specification

  • Open Access

    Deep Learning-Based Reconstruction of Near-Surface Wind Speed Fields from Meteorological Observations

    by Kevin Monsalvez-Pozo, Marcos Martinez-Roig, Nuria P. Plaza-Martín, Cesar Azorin-Molina, Étienne Plésiat

    Reconstruction of near-surface wind speed (NSWS; ~10 m above ground level) from local meteorological station measurements remains an open challenge in climate research. Traditional geostatistical interpolation techniques can provide partial solutions, but their reliability is often limited—especially in regions with complex topography, which are common across Spain. These methods are also highly sensitive to the number of available observations.This work investigates the potential of state-of-the-art Deep Learning (DL) techniques for NSWS reconstruction. In particular, we employ the Climate Reconstruction AI (CRAI) model, an encoder-decoder architecture based on U-Net with partial convolutions, which we train on the Copernicus European Regional Reanalysis (CERRA) dataset, featuring a temporal resolution of three-hourly data and a spatial resolution of 5.5km. This model learns the spatiotemporal patterns of NSWS and is capable of infilling wind fields from grids with missing values.To apply the model to real-world conditions, we focus on reconstructing daily-averaged wind speed fields from incomplete grids of observational data provided by AEMET (the State Meteorological Agency of Spain). We evaluate several model variants to assess the influence of auxiliary variables such as 2-m air temperature, surface air pressure, 2-m relative humidity and orography. In addition to the baseline U-Net approach, a convolutional attention mechanism is employed to capture complex interdependencies among variables — for example, compound events involving relative humidity, temperature, orography and wind such as the Foehn effect — as well as an LSTM-based recurrent module to leverage temporal information in the reconstruction process.The resulting models are benchmarked on CERRA data and subsequently applied to reconstruct NSWS fields from AEMET observations across Spain.


  • Open Access

    EFT analysis of New Physics at COHERENT

    by Víctor Bresó-Pla, Adam Falkowski, Martín González-Alonso, Kevin Monsálvez-Pozo

    Using an effective field theory approach, we study coherent neutrino scattering on nuclei, in the setup pertinent to the COHERENT experiment. We include non-standard effects both in neutrino production and detection, with an arbitrary flavor structure, with all leading Wilson coefficients simultaneously present, and without assuming factorization in flux times cross section. A concise description of the COHERENT event rate is obtained by introducing three generalized weak charges, which can be associated (in a certain sense) to the production and scattering of νe, νμ and $$ {\overline{\nu}}_{\mu } $$ ν ¯ μ on the nuclear target. Our results are presented in a convenient form that can be trivially applied to specific New Physics scenarios. In particular, we find that existing COHERENT measurements provide percent level constraints on two combinations of Wilson coefficients. These constraints have a visible impact on the global SMEFT fit, even in the constrained flavor-blind setup. The improvement, which affects certain 4-fermion LLQQ operators, is significantly more important in a flavor-general SMEFT. Our work shows that COHERENT data should be included in electroweak precision studies from now on.


  • Open Access

    Constraints on leptoquarks from lepton-flavour-violating tau-lepton processes

    by Tomáš Husek, Kevin Monsálvez-Pozo, Jorge Portolés

    Leptoquarks are ubiquitous in several extensions of the Standard Model and seem to be able to accommodate the universality-violation-driven B-meson-decay anomalies and the (g −2)μ discrepancy interpreted as deviations from the Standard Model predictions. In addition, the search for lepton-flavour violation in the charged sector is, at present, a major research program that could also be facilitated by the dynamics generated by leptoquarks. In this article, we consider a rather wide framework of both scalar and vector leptoquarks as the generators of lepton-flavour violation in processes involving the tau lepton. We single out its couplings to leptoquarks, thus breaking universality in the lepton sector, and we integrate out leptoquarks at tree level, generating the corresponding dimension-6 operators of the Standard Model Effective Field Theory. In ref. [1] we obtained model-independent bounds on the Wilson coefficients of those operators contributing to lepton-flavour-violating hadron tau decays and ℓ–τ conversion in nuclei, with ℓ = e, μ. Hence, we use those results to translate the bounds into the couplings of leptoquarks to the Standard Model fermions.


  • Open Access

    Quasi-Dirac neutrinos in the linear seesaw model

    by Carolina Arbeláez, Claudio Dib, Kevin Monsálvez-Pozo, Iván Schmidt

    We implement a minimal linear seesaw model (LSM) for addressing the Quasi-Dirac (QD) behaviour of heavy neutrinos, focusing on the mass regime of $M_{N} \lesssim M_{W}$. Here we show that for relatively low neutrino masses, covering the few GeV range, the same-sign to opposite-sign dilepton ratio, $R_{\ell \ell}$, can be anywhere between 0 and 1, thus signaling a Quasi-Dirac regime. Particular values of $R_{\ell \ell}$ are controlled by the width of the QD neutrino and its mass splitting, the latter being equal to the light-neutrino mass $m_{\nu}$ in the LSM scenario. The current upper bound on $m_{\nu_{1}}$ together with the projected sensitivities of current and future $|U_{N \ell}|^{2}$ experimental measurements, set stringent constraints on our low-scale QD mass regime. Some experimental prospects of testing the model by LHC displaced vertex searches are also discussed.

  • Open Access

    Quasi-Dirac neutrinos in the linear seesaw model

    by Carolina Arbeláez, Claudio Dib, Kevin Monsálvez-Pozo, Iván Schmidt

    We implement a minimal linear seesaw model (LSM) for addressing the Quasi-Dirac (QD) behaviour of heavy neutrinos, focusing on the mass regime of MN ≲ MW. Here we show that for relatively low neutrino masses, covering the few GeV range, the same-sign to opposite-sign dilepton ratio, Rℓℓ, can be anywhere between 0 and 1, thus signaling a Quasi-Dirac regime. Particular values of Rℓℓ are controlled by the width of the QD neutrino and its mass splitting, the latter being equal to the light-neutrino mass mν in the LSM scenario. The current upper bound on mν1 together with the projected sensitivities of current and future |UN ℓ|2 experimental measurements, set stringent constraints on our low-scale QD mass regime. Some experimental prospects of testing the model by LHC displaced vertex searches are also discussed.


  • Open Access

    Lepton-flavour violation in hadronic tau decays and μ-τ conversion in nuclei

    by Tomáš Husek, Kevin Monsálvez-Pozo, Jorge Portolés

    Within the Standard Model Effective Field Theory framework, with operators up to dimension 6, we perform a model-independent analysis of the lepton-flavour-violating processes involving tau leptons. Namely, we study hadronic tau decays and ℓ-τ conversion in nuclei, with ℓ = e, μ. Based on available experimental limits, we establish constraints on the Wilson coefficients of the operators contributing to these processes. Our work paves the way to extract the related information from Belle II and foreseen future experiments.