From 633f94fb6dc7640b3238d61a3e2a870cb98983fe Mon Sep 17 00:00:00 2001 From: showyourwork Date: Mon, 10 Jun 2024 15:36:01 -1000 Subject: [PATCH] [showyourwork] overleaf sync --- src/tex/bib.bib | 53 +++++++++++++++++++++++++++++++++++++++++++++++++ src/tex/ms.tex | 8 ++++---- 2 files changed, 57 insertions(+), 4 deletions(-) diff --git a/src/tex/bib.bib b/src/tex/bib.bib index 4c5c1a4..ba593fe 100644 --- a/src/tex/bib.bib +++ b/src/tex/bib.bib @@ -1,4 +1,24 @@ +@ARTICLE{allred2005, + author = {{Allred}, Joel C. and {Hawley}, Suzanne L. and {Abbett}, William P. and {Carlsson}, Mats}, + title = "{Radiative Hydrodynamic Models of the Optical and Ultraviolet Emission from Solar Flares}", + journal = {\apj}, + keywords = {Methods: Numerical, Radiative Transfer, Sun: Atmosphere, Sun: Flares, Astrophysics}, + year = 2005, + month = sep, + volume = {630}, + number = {1}, + pages = {573-586}, + doi = {10.1086/431751}, +archivePrefix = {arXiv}, + eprint = {astro-ph/0507335}, + primaryClass = {astro-ph}, + adsurl = {https://ui.adsabs.harvard.edu/abs/2005ApJ...630..573A}, + adsnote = {Provided by the SAO/NASA Astrophysics Data System} +} + + + @ARTICLE{asplund2009, author = {{Asplund}, Martin and {Grevesse}, Nicolas and {Sauval}, A. Jacques and {Scott}, Pat}, title = "{The Chemical Composition of the Sun}", @@ -272,6 +292,20 @@ @ARTICLE{doschek2018 adsnote = {Provided by the SAO/NASA Astrophysics Data System} } +@ARTICLE{fisher1984, + author = {{Fisher}, G.~H. and {Canfield}, R.~C. and {McClymont}, A.~N.}, + title = "{Chromospheric evaporation velocities in solar flares}", + journal = {\apjl}, + keywords = {Chromosphere, Evaporation Rate, Solar Flares, Solar X-Rays, Constraints, Hydrodynamics, Solar Physics}, + year = 1984, + month = jun, + volume = {281}, + pages = {L79-L82}, + doi = {10.1086/184290}, + adsurl = {https://ui.adsabs.harvard.edu/abs/1984ApJ...281L..79F}, + adsnote = {Provided by the SAO/NASA Astrophysics Data System} +} + @@ -552,6 +586,25 @@ @ARTICLE{reep2013 adsnote = {Provided by the SAO/NASA Astrophysics Data System} } +@ARTICLE{reep2015, + author = {{Reep}, J.~W. and {Bradshaw}, S.~J. and {Alexander}, D.}, + title = "{Optimal Electron Energies for Driving Chromospheric Evaporation in Solar Flares}", + journal = {\apj}, + keywords = {Sun: chromosphere, Sun: corona, Sun: flares, Astrophysics - Solar and Stellar Astrophysics}, + year = 2015, + month = aug, + volume = {808}, + number = {2}, + eid = {177}, + pages = {177}, + doi = {10.1088/0004-637X/808/2/177}, +archivePrefix = {arXiv}, + eprint = {1506.08115}, + primaryClass = {astro-ph.SR}, + adsurl = {https://ui.adsabs.harvard.edu/abs/2015ApJ...808..177R}, + adsnote = {Provided by the SAO/NASA Astrophysics Data System} +} + @ARTICLE{reep2016, diff --git a/src/tex/ms.tex b/src/tex/ms.tex index 01c72bc..56de15e 100644 --- a/src/tex/ms.tex +++ b/src/tex/ms.tex @@ -57,12 +57,12 @@ \section{Modeling Time-Variable Abundance} We have modified \texttt{ebtel++} to include time-variable abundances as follows. We must first calculate how the abundance factor $f(t)$ changes as flows carry plasma into the corona. Using that, we then incorporate it into the bolometric radiative loss rate, which dominates the cooling for much of a loop's lifetime. The total radiative losses are given by $n^{2} \Lambda(T, n, f)$, where $\Lambda(T, n, f)$ is the radiative loss rate per unit emission measure. To properly calculate the radiative losses, we must update $\Lambda(T, n, f)$ as each of the temperature $T$, density $n$, and abundance factor $f$ changes with time. -In the simulations here, we assume an initial abundance factor $f = 4$ in the corona, meaning that low-FIP elements like iron are enhanced by a factor of 4 over their photospheric values. In comparison, observations of active regions and the solar wind find abundance factors ranging from around $f = 3$ to $f =5$ \citep{brooks2012}, with considerable spatiotemporal variation. When the corona is heated, energy is transported to the chromosphere via thermal conduction, which causes the pressure in the chromosphere to rise and ablate material into the corona (chromospheric evaporation). In the chromosphere, the plasma has photospheric composition with $f = 1.0$, so evaporation causes a shift in the proportion of coronal and photospheric material. The abundance factor $f(t)$ in the corona can then be calculated as a weighted average between the initial material and that carried into the corona via evaporation: +In the simulations here, we assume an initial abundance factor $f = 4$ in the corona, meaning that low-FIP elements like iron are enhanced by a factor of 4 over their photospheric values. In comparison, observations of active regions and the solar wind find abundance factors ranging from around $f = 3$ to $f =5$ \citep{brooks2012}, with considerable spatiotemporal variation. When the corona is heated, energy is transported to the chromosphere via thermal conduction, which causes the pressure in the chromosphere to rise and ablate material into the corona (chromospheric evaporation). In the chromosphere, the plasma has photospheric composition with $f = 1$, so evaporation causes a shift in the proportion of coronal and photospheric material. The abundance factor $f(t)$ in the corona can then be calculated as a weighted average between the initial material and that carried into the corona via evaporation: \begin{align} f(t) &= \frac{f_{0} n_{0} + (n(t) - n_{0})}{n(t)} \nonumber \\ &= 1 + (f_{0} - 1) \frac{n_{0}}{n(t)} \end{align} -\noindent where $n_{0} = n(t=0)$ and $f_{0} = f(t=0)$ are the initial coronal density and abundance factor, and $n(t)$ is the density in the corona as evaporation fills the loop. There is one caveat to this equation: only upflows change the coronal abundance. When the loop is draining, it does not preferentially drain any given elements, so the composition does not change. We therefore only update $f(t)$ when the flows are \textit{into} the corona. +\noindent where $n_{0} = n(t=0)$ and $f_{0} = f(t=0)$ are the initial coronal density and abundance factor, and $n(t)$ is the density in the corona as evaporation fills the loop. There is one caveat to this equation: only upflows change the coronal abundance. \textbf{Unlike the ponderomotive force that likely causes the FIP effect, bulk flows do not cause preferential acceleration of different species. However, since we have assumed that the corona initially has an abundance factor of $f = 4$ and the chromosphere $f = 1$, the evaporation that fills the loop naturally changes the composition. When the loop is draining on the other hand, all elements drain in equal proportion, so the composition of the corona does not change.}\footnote{\textbf{A higher dimensional model might need to update the abundance factor at the top of the chromosphere, as the draining would slightly enhance the proportion of low FIP elements there. However, this is likely to be an exceedingly small effect since the density of the chromosphere is significantly higher than the corona.}} We therefore only update $f(t)$ when the flows are \textit{into} the corona. We then modify the radiative loss function $\Lambda(T, n, f)$, which in previous works has generally only been treated as a function of temperature. In the original implementation of \texttt{ebtel++} (and its IDL-based predecessor \texttt{EBTEL}), the loss function was parameterized as a power-law function in temperature \citep{klimchuk2008}. Here, we have used the v10.1 of the CHIANTI atomic database \citep{dere1997,delzanna2021} to calculate look-up tables of $\Lambda(T, n, f)$ (using discrete steps in $T$, $n$, $f$). Then, at each timestep, we use the current values of $T$, $n$, and $f$ to determine the loss rate with those tables. We use the \citet{asplund2009} dataset for photospheric abundances ($f=1$). As discussed in the appendix of \citet{reep2020}, for other values of $f > 1$, we enhance the abundances of the low-FIP elements (\textit{e.g.} for $f=4$, the abundance of iron is 4 times greater than in the \citealt{asplund2009} dataset). @@ -92,7 +92,7 @@ \section{Loop Simulations} It is clear that the heating rates, and thus the strength of chromospheric evaporation, play a large role in the evolution of the abundance factor and therefore the radiative losses. Since radiative losses are stronger with coronal abundances than with photospheric ones, this causes the time-variable case to cool faster than the photospheric case but slower than the coronal case in general. In the weakest heating case, the coronal density increases by around 50\%, causing the abundance factor to fall to 3 over the course of 10 minutes, compared to the strongest heating case where the abundance factor falls to photospheric levels almost immediately. Notably, since radiation is relatively weak prior to the end of chromospheric evaporation, the temperatures and densities of the impulsive phase are relatively unaffected by the abundance values. The cases only start to diverge in the cooling period, after radiation becomes the dominant cooling mechanism. We note that short heating pulses in \texttt{ebtel++} (like those used here) overestimate the density relative to higher dimensional models by around 20\% because of the lack of spatial extent \citep{barnes2016}, which may exacerbate the divergences. -In Figure \ref{fig:L80}, we show a similar comparison with 80 Mm loops. Since the cooling time depends on loop length ($\propto L^{5/6}$, \citealt{cargill1995}), we expect that the differences should be exaggerated here. While the overall trends are similar to the previous case, there are a few differences worth noting. The first is that the coronal density does not grow as large with the same heating rates, and as a consequence the abundance factor remains somewhat higher. The second is that the duration of evaporation is somewhat longer, as the flows must travel a longer distance to fill the loop, so the rate of change of the abundance factor is generally slower. +In Figure \ref{fig:L80}, we show a similar comparison with 80 Mm loops. Since the cooling time depends on loop length ($\propto L^{5/6}$, \citealt{cargill1995}), we expect that the differences should be exaggerated here. While the overall trends are similar to the previous case, there are a few differences worth noting. The first \textbf{difference} is that the coronal density does not grow as large with the same heating rates, and as a consequence the abundance factor remains somewhat higher. \textbf{This is partially a consequence of the assumed form of heating, which in \texttt{ebtel++} is equivalent to a thermal conduction front. Different heating mechanisms, such as non-thermal electron beams, would deposit energy directly in the chromosphere, which affects the flows and thus the resultant coronal densities \citep{fisher1984,allred2005,reep2015}.} The second \textbf{difference} is that the duration of evaporation is somewhat longer, as the flows must travel a longer distance to fill the loop, so the rate of change of the abundance factor is generally slower. \begin{figure*} \script{render_figure2.py} \centering @@ -140,7 +140,7 @@ \section{Discussion} \section*{Acknowledgments} \textbf{All of the tools used to produce this paper are open-source.} At the time of writing, the implementation into \texttt{ebtel++} is publicly available as an unmerged pull request on GitHub: \url{https://github.com/rice-solar-physics/ebtelPlusPlus/pull/87}. -This paper is fully reproducible with the \href{https://github.com/showyourwork/showyourwork}{\showyourwork} package, and can be downloaded from \url{https://github.com/jwreep/ebtel_abundances}. \textbf{This package strives for full reproducibility and transparency in science. Each figure in this paper contains a hyperlink with the Github logo that points to the script used to generate the plots in that figure. Additionally, the whole paper can be generated with a single command (`showyourwork build`) that will run the \texttt{ebtel++} simulations, read in the data, and produce the figures. Users can then additionally reconfigure the simulations, the plots, or the scripts to their own liking to validate or extend this research.} +This paper is fully reproducible with the \href{https://github.com/showyourwork/showyourwork}{\showyourwork} package, and can be downloaded from \url{https://github.com/jwreep/ebtel_abundances}. \textbf{This package strives for full reproducibility and transparency in science. Each figure in this paper contains a hyperlink with the Github logo that points to the script used to generate the plots in that figure. Additionally, the whole paper can be generated with a single command (`showyourwork build') that will run the \texttt{ebtel++} simulations, read in the data, and produce the figures. Users can then additionally reconfigure the simulations, the plots, or the scripts to their own liking to validate or extend this research.} \software{ \texttt{ebtel++} (\url{https://github.com/rice-solar-physics/ebtelPlusPlus/pull/87}),