Pasar al contenido principal

APRENDIZAJE AUTOMATICO PARA HIDROGEOLOGOS FORENSES

PRE2020-093145

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal de Promoción del Talento y su Empleabilidad en I+D+i
Subprograma Subprograma Estatal de Formación
Convocatoria Contratos Predoctorales
Año convocatoria 2020
Plan Estatal 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSITAT POLITÈCNICA DE VALÈNCIA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 4
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Ensemble random forest filter: An alternative to the ensemble Kalman filter for inverse modeling

  • Napa-García, Gian F.
  • Gómez-Hernández, J. Jaime|||0000-0002-0720-2196
  • Godoy, Vanessa A.|||0000-0002-2594-7351
[EN] The ensemble random forest filter (ERFF) is presented as an alternative to the ensemble Kalman filter (EnKF) for inverse modeling. The EnKF is a data assimilation approach that forecasts and updates parameter estimates sequentially in time as observations are collected. The updating step is based on the experimental covariances computed from an ensemble of realizations, and the updates are given as linear combinations of the differences between observations and forecasted system state values. The ERFF replaces the linear combination in the update step with a non-linear function represented by a random forest. This way, the non-linear relationships between the parameters to be updated and the observations can be captured, and a better update produced. The ERFF is demonstrated for log-conductivity identification from piezometric head observations in several scenarios with varying degrees of heterogeneity (log-conductivity variances going from 1 up to 6.25 (ln m/d)2), number of realizations in the ensemble (50 or 100), and number of piezometric head observations (18 or 36). In all scenarios, the ERFF works well, reconstructing the log-conductivity spatial heterogeneity while matching the observed piezometric heads at selected control points. For benchmarking purposes, the ERFF is compared to the restart EnKF to find that the ERFF is superior to the EnKF for the number of ensemble realizations used (small in typical EnKF applications). Only when the number of realizations grows to 500 the restart EnKF can match the performance of the ERFF, albeit at more than double the computational cost., The authors acknowledge grant PID2019-109131RB-I00 funded by MCIN/AEI/10.13039/501100011033 and project InTheMED, which is part of the PRIMA Programme supported by the European Union's Horizon 2020 Research and Innovation Programme under Grant Agreement No 1923.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

genES-MDA: a generic open-source software package to solve inverse problems via the Ensemble Smoother with Multiple Data Assimilation

  • Todaro, Valeria
  • D'Oria, Marco
  • Tanda, Maria Giovanna
  • Gómez-Hernández, J. Jaime|||0000-0002-0720-2196
[EN] Ensemble Kalman filter methods have been successfully applied for data assimilation and parameter estimation through inverse modeling in various scientific fields. We have developed a new generic software package for the solution of inverse problems implementing the Ensemble Smoother with Multiple Data Assimilation (genES-MDA). It is an open-source, platform-independent Python-based program. Its aim is to facilitate the management and configuration of the ES-MDA through several programming tools that help in the preparation of the different steps of ES-MDA. genES-MDA has a flexible workflow that can be easily adapted for the implementation of different variants of the ensemble Kalman filter and for the solution of generic inverse problems. This paper presents a description of the package and some application examples. genES-MDA has been tested in three synthetic case studies: the solution of the reverse flow routing for the estimation of the inflow hydrograph to a river reach using observed water levels and a calibrated forward model of the river system, the identification of a hydraulic conductivity field using piezometric observations and a known forward flow model, and the estimation of the release history of a contaminant spill in an aquifer from measured concentration data and a known flow and transport model. The results of all these tests have demonstrated the flexibility of genES-MDA and its capabilities to efficiently solve different types of inverse problems., The fourth author would like to acknowledge grant PID2019-109131RB-I00 funded by the Spanish MCIN/AEI/10.13039/501100011033 and the University of Parma for hosting him as a visiting professor, during which time the paper was finalized. The authors are thankful to the anonymous Reviewers for their valuable comments.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Simultaneous identification of a non-point contaminant source with Gaussian spatially distributed release and heterogeneous hydraulic conductivity in an aquifer using the LES-MDA method

  • Zhang, Wenjun
  • Xu, Teng
  • Chen, Zi
  • Lu, Chunhui
  • Yang, Jie
  • Ye, Yu
  • Miao Jing
  • Gómez-Hernández, J. Jaime|||0000-0002-0720-2196
[EN] Space temporal distribution of the contaminant plumes and aquifer properties is critical for groundwater management. However, most previous studies have focused on point source identification, barely exploring the identification of non -point sources. Xu et al. (2022) proposed to identify non -point sources but did not consider uncertainties in aquifer properties and release mass loading. In this work, we have implemented an application of the localized ensemble smoother with multiple data assimilation (LES-MDA) for the simultaneous identification of Gaussian hydraulic conductivities and non -point source parameters including Gaussian release mass -loading by assimilating both piezometric head and concentration observations in a synthetic confined aquifer. The results prove that the LES-MDA is not only capable of providing accurate identification of the spatial architecture of non -point contaminant sources and related release parameters (such as initial release time, and release duration) but also spatially heterogeneous release mass -loading and hydraulic conductivities., Financial support to carry out this work was received from the financial support from the National Key Research and Development Project (2021YFC3200500) , and the National Natural Science Foundation of China (Grant No. 42377046) . J.J. Gomez-Hernandez acknowledges the financial support funded by MCIN/AEI/10.13039/50110001103 through grant PID2019-109131RB-I00. C. Lu acknowledges the National Natural Science Foundation of China (51879088) , Fundamental Research Funds for the Central Universities (B200204002) , and the Natural Science Foundation of Jiangsu Province (BK 20190023) .
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Predicting discharge from a complex karst system using the ensemble smoother with multiple data assimilation

  • Pansa, Alessandro
  • Butera, Ilaria
  • Vigna, Bartolomeo
  • Gómez-Hernández, J. Jaime|||0000-0002-0720-2196
[EN] Can the ensemble smoother with multiple data assimilation be used to predict discharge in an Alpine karst aquifer? The answer is yes, at least, for the Bossea aquifer studied. The ensemble smoother is used to fit a unit hydrograph simultaneously with other parameters in a hydrologic model, such as base flow, infiltration coefficient, or snow melting contribution. The fitting uses observed discharge flow rates, daily precipitations, and temperatures to define the model parameters. The data assimilation approach gives excellent results for fitting individual events. After the analysis of 27 such events, two average models are defined to be used to predict flow discharge from precipitation and temperature, one model for prediction during spring (when snow melting has an impact) and another one during autumn, yielding acceptable results, particularly for the fall rainfall events. The lesser performance for the spring events may indicate that the snow melting approximation needs to be revised. The results also show that the parameterization of the infiltration coefficient needs further exploration. Overall, the main conclusion is that the ensemble smoother could be used to define a characteristic "signature" of a karst aquifer to be used in forecast analyses. The reasons for using the ensemble smoother instead of other stochastic approaches are that it is easy to use and explain and provides an estimation of the uncertainty about the predictions., J. Jaime Gomez-Hernandez acknowledges grant PID2019109131RB-I00 funded by MCIN/AEI/10.13039/501100011033.