[Loops] FW: summaries on nanoflare debates in "coronalloopworkshops"

Thu Mar 12 13:59:53 MDT 2009

Hi Jim, Robertus and others,

I left the nanoflare business some time ago. I will thus reply more at  
the philosophical level rather than on the technical details.

The one thing that I remember from my own work is that nanoflare  
heating is an absolutely hopeless business. It is based on the  
assumption that we can extrapolate (!) the slope of the distribution  
over many order of magnitude (!) into unobserved territory (!). It  
only works if you take as extrapolating slope the few outlier studies  
(!) that find a slope steep enough. And you must believe (!) that  
sufficiently small events can actually really exist.

There was a time when I was ready to take all these hurdles. What  
really killed my faith however is the realization that, when analyzing  
data and identifying members of the flare distribution, one needs to  
identify flares one by one. At small scales (where most events are)  
you essentially see a bunch of brightening pixels and one needs to  
group these pixels in events. The 'grouping criterium' can be strict  
('touching' pixels belong to the same event) or somewhat more relaxed  
('close-by' pixels can belong to the same event). In the first case  
you get many small events, in the last case you get fewer & bigger  
events.  Comparing the two cases, will give you widely different slopes.

My conclusion: determining the slope for gradually smaller flares is  
gradually more subjective. We are not proving anything by measuring &  
extrapolating slopes of small event distributions.

I do not mean to say that nanoflares (whatever small energy releases  
at unobserved scales) are not interesting or not relevant for coronal  
heating. At the contrary, I am personally looking forward to see their  
appearance in the so-called hot channel of the high-res EUV imager  
foreseen on Solar Orbiter! I am just saying that extrapolating ill- 
defined power laws is, in my humble opinion, not proving anything.

David

On 11 Mar 2009, at 14:55, Klimchuk, James A. (GSFC-6710) wrote:

> Hi Robertus,
>
>  Thanks for sending David's presentation.   He heard about our
> listserve discussion and sent me the paper that goes with it.  I in  
> turn
> sent him the following two messages and am waiting for his reply.
>
> Cheers,
> Jim
>
>
> Hi David,
>
>  Yes, we've been having an e-mail discussion within the loops
> community, and Robertus mentioned that you had ruled out nanoflare
> heating.  Thanks for the paper and presentation.  I've only had a  
> chance
> to glance at them, but they look very interesting.
>  Do you stand by your claim?  As you know, it is extremely difficult
> to measure event energies, especially for small events.  I've had some
> discussions with Arnold Benz recently, and he feels strongly that we  
> do
> not have accurate enough information to place reliable constraints on
> the problem, i.e., to determine whether nanoflares are important based
> on energy distributions.  Clare Parnell tells me the same thing.   
> Also,
> I would note that recent studies by the Max Planck group of small  
> events
> observed by SUMER give power laws steeper than 2, supporting the
> nanoflare idea.
>  Nanoflares of the type I believe exist are so small that they cannot
> be measured directly, and there is a danger in extrapolating  
> "observed"
> energy distributions to lower values.  The physics of large eruptive
> flares is probably very different from the physics of small confined
> flares and nanoflares.  A recent study by Yashiro et al. finds that
> flares that are not associated with CMEs have a power law steeper than
> 2.
>  I'm very interested in your feedback.
>
> Best wishes,
> Jim
>
>
> Hi again,
>
>   I was just looking at Section 5.1 of your paper (slide 29 in the
> presentation), where you discuss lower limits to the nanoflare energy
> based on the arguments of Markus (1999).  I'm afraid I see some  
> problems
> with those arguments.
>
>   First, Markus' scaling laws are P~T^3 and L~T.  This gives
> E_th~PL~T^4, not T^6.
>
>   Second, the coronal pressure is not set by the chromospheric
> pressure.  Rather, the chomosphere will extend to a height at which  
> its
> gravitationally stratified pressure matches that of the corona,  
> which is
> set by the heating.  Nonetheless, I agree that there is practical  
> lower
> limit to the coronal pressure produced by a nanoflare.
>
>   Third, the cross sections of the strands that are heated by
> nanoflares are probably very small.  For example, a typical TRACE loop
> contains several tens of elemental flux tubes as observed in the
> photosphere.  If energy release occurs at the interfaces of these  
> tubes,
> as seems very likely, then the nanoflare-heated strands could have a
> diameter of 10 km or less.
>
>   The bottom line is that 10^23 erg is not a valid lower limit to the
> energy of nanoflares.
>
> Cheers,
> Jim
>
>
>
> -----Original Message-----
> From: Robert von Fay-Siebenburgen [mailto:robertus at sheffield.ac.uk]
> Sent: Monday, March 09, 2009 4:16 PM
> To: Klimchuk, James A. (GSFC-6710)
> Subject: Re: [Loops] summaries on nanoflare debates in
> "coronalloopworkshops"
>
> jim,
>
> see attached:),
> cheers,
> r.
>
> On Mon, 9 Mar 2009, Klimchuk, James A. (GSFC-6710) wrote:
>
>> Valery and Robert,
>>
>>
>>
>> Thanks for bringing up waves.  I have two quick points.
>>
>>
>>
>> 1.  The internal plasma structure of a loop (multi-stranded or not)  
>> is
>> determined by the heating and cannot be decoupled from the wave
>> dissipation (unless the wave is energetically insignificant).  See
>> Section 4.3 of the attached.
>>
>>
>>
>> 2.  There may be different opinions of what is meant by "nanoflare."
>> Some people define it to be an identifiable point-like brightening
> that
>> probably occurs in a tiny bipole.  Most loops modelers define it to  
>> be
>> an impulsive energy release in a long and unresolved magnetic strand.
>> It cannot be directly observed.  With this definition, even resonant
>> wave absorption produces nanoflare heating (also explained in  
>> Sections
>> 4.3 and 4.2).  I'm very curious about David Berghmans' talk and have
> my
>> doubts about the conclusions!
>>
>>
>>
>> Cheers,
>>
>> Jim
>>
>>
>>
>> ________________________________
>>
>> From: loops-bounces at solar.physics.montana.edu
>> [mailto:loops-bounces at solar.physics.montana.edu] On Behalf Of
>> Nakariakov, Valery
>> Sent: Saturday, March 07, 2009 9:02 AM
>> To: A mailing list for scientists involved in the observation and
>> modeling ofsolar loop structures; A mailing list for scientists
> involved
>> in the observation and modelingofsolar loop structures
>> Subject: Re: [Loops] summaries on nanoflare debates in
>> "coronalloopworkshops"
>>
>>
>>
>> Hi Robertus, hi all,
>>
>>> however, both hinode and very recent rosa observations have shown  
>>> the
>> omnipotent presence of mhd waves ... that have the necessary poyting
>> flux to heat the lower
>> atmosphere or corona. so, it is not just reconnection!
>>
>> There are two important aspects, in my opinion:
>> (a) even if the Poynting flux is sufficient, there also should be
>> sufficiently effective mechanisms for the wave energy dissipation  
>> at a
>> reasonable height. The efficiency of both phase mixing and RA is
>> determined by the steepness of the transverse structuring of the
>> waveguiding plasma structures.
>>
>> Structuring is definitely an unknown parameter, as you've pointed  
>> out:
>>> if (IF) we want to get closer to loop fine structure
>> (multi-thread vs whatever) or uncover the operating heating function,
> a
>> very promising and currently already available way is of the one by
>> magneto-seismology.
>>
>> I agree. As I have mentioned in this forum before, there are several
>> coronal seismological indications of the subresolution structuring of
>> the corona. But, what is important is not only the spatial scale of
>> structuring, but also the gradients of the plasma parameters.
>>
>> (b) The amplitudes of the observed transverse waves are sufficiently
>> high to produce a significant change of the magnetic field in the
>> vicinity of the potential reconnection
>> sites. Hence, the waves can effectively contribute to reconnection,
> and
>> then the separation of "wave-based" and "reconnection-based"
> mechanisms
>> for heating seems to be a bit too artificial.
>>
>> All the best,
>>
>> Valery
>>
>>
>>
>>
> <Prague.pdf>_______________________________________________
> Loops mailing list
> Loops at solar.physics.montana.edu
> https://mithra.physics.montana.edu/mailman/listinfo/loops

---
David Berghmans
Royal Observatory of Belgium