Inconsistency in temperatures used to calculate column densities in beam heating?

[jwreep]

The electron temperature is used to determine whether the a cell is below OPTICALLY_THICK_TEMPERATURE when calculating column densities:

HYDRAD/HYDRAD/source/eqns.cpp

Lines 2116 to 2128 in 0fa1f99

	#ifdef OPTICALLY_THICK_RADIATION
	if( CellProperties.T[ELECTRON] < OPTICALLY_THICK_TEMPERATURE )
	{
	// Calculate the neutral hydrogen column number density
	fHI_c += ( CellProperties.HI * CellProperties.n[HYDROGEN] ) * CellProperties.cell_width;
	CellProperties.HI_c = fHI_c;
	#ifndef NLTE_CHROMOSPHERE
	// Calculate the mass column density
	frho_c += CellProperties.rho[1] * CellProperties.cell_width;
	CellProperties.rho_c = frho_c;
	#endif // NLTE_CHROMOSPHERE
	}
	#endif // OPTICALLY_THICK_RADIATION

but the LTE chromosphere uses the hydrogen temperature, and even has a comment about using consistent rates:

HYDRAD/HYDRAD/source/eqns.cpp

Lines 1160 to 1178 in 0fa1f99

	/////////////////////////////////////////////////////////////////
	// Calculate the electron density for the LTE chromosphere //
	/////////////////////////////////////////////////////////////////
	// NOTE: We can't use the non-equilibrium population here because it's based on a completely different set of rates.
	// If you try to use it (or the RadiationModel object to get an equilibrium population) for CellProperties.HI
	// then you end up with a completely different value for the neutral hydrogen population at T[HYDROGEN] than
	// pHI->GetIonFrac returns and, consequently, a different electron density. The non-equilibrium populations quickly
	// evolves to a new equilibrium in the dense lower atmosphere. The different electron density leads to a different
	// electron temperature (very different to the hydrogen temperature) in the lower chromosphere, particularly at the
	// boundaries which don't get updated except by refinement and quickly propagates into the domain, messing
	// everything up quite spectacularly!
	// The bottom line is: USE CONSISTENT RATES AND ION POPULATIONS EVERYWHERE
	if( CellProperties.T[HYDROGEN] < OPTICALLY_THICK_TEMPERATURE )
	CellProperties.HI = pHI->GetIonFrac( log10(CellProperties.T[HYDROGEN]) );
	else
	CellProperties.HI = 0.0;

Perhaps one could argue that the temperatures should be quickly collisionally equilibrated, but I think we could sidestep any potential issues by changing this line to use the hydrogen temperature:

HYDRAD/HYDRAD/source/eqns.cpp

Line 2117 in 0fa1f99

if( CellProperties.T[ELECTRON] < OPTICALLY_THICK_TEMPERATURE )

Or am I missing something?

[jwreep]

I'm guess I'm missing that the NLTE chromosphere seems to use the electron temperature to calculate excitation rates, so perhaps we might want to check all of this for consistency.

[sjbradshaw]

This is a difficult one because the electron density is needed in order to calculate the electron temperature from the electron energy. To do this, we first need the hydrogen population fractions, from which we can find the electron density (and add a small correction due to ionized trace elements in LTE), but of course the population fractions are a function of the electron temperature… So, there are two choices: 1. Use the proton/hydrogen temperature (assuming Ti ~ Te at high density), which we have ready access to; or 2. Use an initial guess for the electron temperature to find the hydrogen population fractions and then recalculate Te with the new electron density (or recalculate the electron energy with the new electron density, to keep Te the same). I remember spending quite a while on this issue before settling on the way it is currently handled as a compromise. However, if it’s of concern then I recommend doing a few experiments along the lines of the above (including changing the switch between optically-thin / thick radiation to the proton/hydrogen temperature) to see what the differences are. From: Jeffrey Reep ***@***.***> Sent: Thursday, March 14, 2024 10:12 PM To: rice-solar-physics/HYDRAD ***@***.***> Cc: Subscribed ***@***.***> Subject: [rice-solar-physics/HYDRAD] Inconsistency in temperatures used to calculate column densities in beam heating? (Issue #104) The electron temperature is used to determine whether the a cell is below OPTICALLY_THICK_TEMPERATURE when calculating column densities: https://github.com/rice-solar-physics/HYDRAD/blob/0fa1f9949540e6b9379548c72a030621fdee381d/HYDRAD/source/eqns.cpp#L2116-L2128 but the LTE chromosphere uses the hydrogen temperature, and even has a comment about using consistent rates: https://github.com/rice-solar-physics/HYDRAD/blob/0fa1f9949540e6b9379548c72a030621fdee381d/HYDRAD/source/eqns.cpp#L1160-L1178 Perhaps one could argue that the temperatures should be quickly collisionally equilibrated, but I think we could sidestep any potential issues by changing this line to use the hydrogen temperature: https://github.com/rice-solar-physics/HYDRAD/blob/0fa1f9949540e6b9379548c72a030621fdee381d/HYDRAD/source/eqns.cpp#L2117 Or am I missing something? — Reply to this email directly, view it on GitHub <#104> , or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACC6C7VJJRVRFKXNPRSJRZLYYJRHPAVCNFSM6AAAAABEXIBDHKVHI2DSMVQWIX3LMV43ASLTON2WKOZSGE4DONRTGQZDKNI> . You are receiving this because you are subscribed to this thread. <https://github.com/notifications/beacon/ACC6C7UX4PEXCLP4FPTY4F3YYJRHPA5CNFSM6AAAAABEXIBDHKWGG33NNVSW45C7OR4XAZNFJFZXG5LFVJRW63LNMVXHIX3JMTHIEZFGJ4.gif> Message ID: ***@***.*** ***@***.***> >

[jwreep]

I'll do some simple tests. It might be completely unimportant, but it caught my attention this week. I'm looking at implementing some return current heating, so reviewing what is currently in there.

[sjbradshaw]

Fair enough – at some point I meant to get around to implementing the chromospheric losses for flares too (Hong, Carlsson, Ding, 2022, A&A 661, A77). From: Jeffrey Reep ***@***.***> Sent: Friday, March 15, 2024 5:31 PM To: rice-solar-physics/HYDRAD ***@***.***> Cc: sjbradshaw ***@***.***>; Comment ***@***.***> Subject: Re: [rice-solar-physics/HYDRAD] Inconsistency in temperatures used to calculate column densities in beam heating? (Issue #104) I'll do some simple tests. It might be completely unimportant, but it caught my attention this week. I'm looking at implementing some return current heating, so reviewing what is currently in there. — Reply to this email directly, view it on GitHub <#104 (comment)> , or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACC6C7WKK4A3DYRWPHLTXCTYYNZBRAVCNFSM6AAAAABEXIBDHKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAMBQGU4DKOJUGQ> . You are receiving this because you commented. <https://github.com/notifications/beacon/ACC6C7WMPTK2M2WEBNMRACTYYNZBRA5CNFSM6AAAAABEXIBDHKWGG33NNVSW45C7OR4XAZNMJFZXG5LFINXW23LFNZ2KUY3PNVWWK3TUL5UWJTTXH2CNQ.gif> Message ID: ***@***.*** ***@***.***> >