From 01c8dd71c3b9c60e4626d3fded09b27e1bcbf873 Mon Sep 17 00:00:00 2001 From: Matt Borland Date: Tue, 1 Sep 2026 11:21:45 -0400 Subject: [PATCH 1/2] [JOSS Review] Mapping f not rendering The previously used f did not render correctly, so I have inserted just an ascii lowercase f --- paper/paper.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/paper/paper.md b/paper/paper.md index 24a1421..e91b859 100755 --- a/paper/paper.md +++ b/paper/paper.md @@ -37,7 +37,7 @@ We present a Knowledge-Informed Mapping toolkit in Python programming language, KIM offers a preliminary understanding of data interdependencies while optimizing the training step with uncertainty accounted for. We expect this toolkit will be helpful to glue the model data integration for Earth science applications. # Statement of need -Striving for scientific hypothesis testing and discovery, Earth scientists oftentimes develop data-driven mappings -- either for inverse modeling, as part of model calibration, or forward modeling, as an emulator. Both approaches benefit from an efficient way of mapping, $ƒ$, that projects from a vector of inputs $\mathbf{X}$ to a vector of outputs $\mathbf{Y}$. +Striving for scientific hypothesis testing and discovery, Earth scientists oftentimes develop data-driven mappings -- either for inverse modeling, as part of model calibration, or forward modeling, as an emulator. Both approaches benefit from an efficient way of mapping, $f$, that projects from a vector of inputs $\mathbf{X}$ to a vector of outputs $\mathbf{Y}$. Such mapping approach has seen successes in addressing inverse and forward problems in multiple studies across Earth sciences [@Krasnopolsky:2003; @HU:2014; @Cromwell:2021; @Mudunuru:2022]. Nevertheless, constructing the mapping $ƒ$ that connects all inputs $\mathbf{X}$ to all outputs $\mathbf{Y}$ is usually challenging due to (1) limited data/simulations for training; (2) uninformative relations between some members of $\mathbf{X}$ and $\mathbf{Y}$; and (3) the structural uncertainty of the mapping $ƒ$. To that, @Jiang:2023 and @Wang:2025 leveraged the idea of integrating scientific knowledge with deep learning [@Willard:2022] to develop knowledge-informed mapping (KIM). The goal of this paper is to document and open source KIM for a general public usage. \autoref{fig:kim} shows the general procedures of KIM which are detailed in the next section. @@ -87,4 +87,4 @@ We present two applications of KIM in performing inverse modeling, with Jupyter This work was supported by both the Laboratory Directed Research and Development Program at Pacific Northwest National Laboratory and the IDEAS-Watersheds project. The Laboratory Directed Research and Development Program at Pacific Northwest National Laboratory is a multiprogram national laboratory operated by Battelle for the U.S. Department of Energy. Pacific Northwest National Laboratory is operated for the DOE by Battelle Memorial Institute under contract DE-AC05-76RL01830. The IDEAS-Watersheds project is funded by the U.S. Department of Energy (DOE), Office of Science (SC) Biological and Environmental Research (BER) program, as part of BER’s Environmental System Science (ESS) program. -# References \ No newline at end of file +# References From e848655b393397fde844d39ea221cd7dd2fe60b0 Mon Sep 17 00:00:00 2001 From: Matt Borland Date: Tue, 1 Sep 2026 11:23:45 -0400 Subject: [PATCH 2/2] Additional instances of non-rendered mapping f --- paper/paper.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/paper/paper.md b/paper/paper.md index e91b859..d04dc29 100755 --- a/paper/paper.md +++ b/paper/paper.md @@ -33,14 +33,14 @@ bibliography: paper.bib \textsuperscript{2} Civil, Construction and Environmental Engineering, University of Alabama, Tuscaloosa, AL, USA # Summary -We present a Knowledge-Informed Mapping toolkit in Python programming language, named KIM, to optimize the development of the mapping $ƒ$ from a vector of inputs $\mathbf{X}$ to a vector of outputs $\mathbf{Y}$. KIM builds on the methodology development of deep learning-based inverse mapping in @Jiang:2023 and @Wang:2025. +We present a Knowledge-Informed Mapping toolkit in Python programming language, named KIM, to optimize the development of the mapping $f$ from a vector of inputs $\mathbf{X}$ to a vector of outputs $\mathbf{Y}$. KIM builds on the methodology development of deep learning-based inverse mapping in @Jiang:2023 and @Wang:2025. KIM offers a preliminary understanding of data interdependencies while optimizing the training step with uncertainty accounted for. We expect this toolkit will be helpful to glue the model data integration for Earth science applications. # Statement of need Striving for scientific hypothesis testing and discovery, Earth scientists oftentimes develop data-driven mappings -- either for inverse modeling, as part of model calibration, or forward modeling, as an emulator. Both approaches benefit from an efficient way of mapping, $f$, that projects from a vector of inputs $\mathbf{X}$ to a vector of outputs $\mathbf{Y}$. Such mapping approach has seen successes in addressing inverse and forward problems in multiple studies across Earth sciences [@Krasnopolsky:2003; @HU:2014; @Cromwell:2021; @Mudunuru:2022]. -Nevertheless, constructing the mapping $ƒ$ that connects all inputs $\mathbf{X}$ to all outputs $\mathbf{Y}$ is usually challenging due to (1) limited data/simulations for training; (2) uninformative relations between some members of $\mathbf{X}$ and $\mathbf{Y}$; and (3) the structural uncertainty of the mapping $ƒ$. To that, @Jiang:2023 and @Wang:2025 leveraged the idea of integrating scientific knowledge with deep learning [@Willard:2022] to develop knowledge-informed mapping (KIM). The goal of this paper is to document and open source KIM for a general public usage. \autoref{fig:kim} shows the general procedures of KIM which are detailed in the next section. +Nevertheless, constructing the mapping $f$ that connects all inputs $\mathbf{X}$ to all outputs $\mathbf{Y}$ is usually challenging due to (1) limited data/simulations for training; (2) uninformative relations between some members of $\mathbf{X}$ and $\mathbf{Y}$; and (3) the structural uncertainty of the mapping $f$. To that, @Jiang:2023 and @Wang:2025 leveraged the idea of integrating scientific knowledge with deep learning [@Willard:2022] to develop knowledge-informed mapping (KIM). The goal of this paper is to document and open source KIM for a general public usage. \autoref{fig:kim} shows the general procedures of KIM which are detailed in the next section. ![Comparison between KIM and the original mapping.\label{fig:kim}](../docs/figures/Figure-KIM.png){ width=80% }