arXiv:2511.16926v1 Announce Type: cross 
Abstract: We propose the Diffusion-Inversion-Net (DIN) framework for inverse modeling of groundwater flow and solute transport processes. DIN utilizes an offline-trained Denoising Diffusion Probabilistic Model (DDPM) as a powerful prior leaner, which flexibly incorporates sparse, multi-source observational data, including hydraulic head, solute concentration, and hard conductivity data, through conditional injection mechanisms. These conditioning inputs subsequently guide the generative inversion process during sampling. Bypassing iterative forward simulations, DIN leverages stochastic sampling and probabilistic modeling mechanisms to directly generate ensembles of posterior parameter fields by repeatedly executing the reverse denoising process. Two representative posterior scenarios, Gaussian and non-Gaussian, are investigated. The results demonstrate that DIN can produce multiple constraint-satisfying realizations under identical observational conditions, accurately estimate hydraulic-conductivity fields, and achieve reliable uncertainty quantification. The framework exhibits strong generalization capability across diverse data distributions, offering a robust and unified alternative to conventional multi-stage inversion methodologies.

تم تقديم إطار عمل Diffusion-Inversion-Net (DIN) كنهج مبتكر لنمذجة التدفقات الجوفية وعملية نقل المذاب، باستخدام نموذج احتمالي للتنقيح (DDPM) لدمج بيانات المراقبة المتنوعة. يسمح هذا الإطار بتوليد حقول المعلمات اللاحقة دون الحاجة إلى محاكاة أمامية تكرارية، مما يعزز الكفاءة في النمذجة.

Se ha presentado el marco Diffusion-Inversion-Net (DIN) como un enfoque novedoso para la modelización inversa de los flujos de agua subterránea y los procesos de transporte de solutos, utilizando un Modelo Probabilístico de Difusión Denoising (DDPM) para incorporar diversos datos de observación. Este marco permite la generación de campos de parámetros posteriores sin necesidad de simulaciones hacia adelante iterativas, mejorando así la eficiencia en la modelización.

Le cadre Diffusion-Inversion-Net (DIN) a été introduit comme une nouvelle approche pour la modélisation inverse des flux d'eau souterraine et des processus de transport de solutés, utilisant un Modèle Probabiliste de Diffusion Denoising (DDPM) pour intégrer diverses données d'observation. Ce cadre permet la génération de champs de paramètres postérieurs sans avoir besoin de simulations avant itératives, améliorant ainsi l'efficacité de la modélisation.

The Diffusion-Inversion-Net (DIN) framework has been introduced as a novel approach for inverse modeling of groundwater flow and solute transport processes, utilizing a Denoising Diffusion Probabilistic Model (DDPM) to incorporate various observational data. This framework allows for the generation of posterior parameter fields without the need for iterative forward simulations, enhancing efficiency in modeling.

Diffusion-Inversion-Net (DIN): An End-to-End Direct Probabilistic Framework for Characterizing Hydraulic Conductivities and Quantifying Uncertainty

Was this article worth reading? Share it

Attentive AI

DataCircle.io

DEIbias.io