experiments on increase critical colatitude limits for use of Heelis potential in dynamo

[ZhuxiaoLi66]

Aiming to change the TEC distribution along low- midlatitude (make it more centralise toward equator) with changing the ion/electron circulation, Tim & Astrid designed a test to increase the critical colatitude limits for use of Heelis potential in dynamo.
For the detailed IPE code, increase the values of offset1_deg, offset2_deg by 5 deg to change crit(1) & crit(2) in module_colath.f in ./dynamo, so that to change pfrac (fraction of dynamo in solution in the NH. =1 low lat, =0 hi lat).
The runs for two storms (20230420 & 20031120) has been carried out, the results are encouraging and as expected with comparison with the control runs by the WAM-IPE tip version.
For record, I revised the namelist varible values in GSMWAM-IPE/scripts/compsets/exglobal/exglobal_fcst_nems.sh to change the values of offset1_deg, offset2_deg.

[ZhuxiaoLi66]

The following plots are the TEC comparison results between the test run and control run for 20230424 storm.

[ZhuxiaoLi66]

The following plots the same but for 20031120 and 20031121 two days.

[timfullerrowell]

Thanks Zhuxiao, Yes, these look promising, as you say. Be interesting to see the bias numbers for the recent storm against Glo-TEC. Tim

…

On Thu, Sep 28, 2023 at 11:15 AM ZhuxiaoLi ***@***.***> wrote: Aiming to change the TEC distribution along low- midlatitude (make it more centralise toward equator) with changing the ion/electron circulation, Tim & Astrid designed a test to increase the critical colatitude limits for use of Heelis potential in dynamo. For the detailed IPE code, increase the values of offset1_deg, offset2_deg by 5 deg to change crit(1) & crit(2) in module_colath.f in ./dynamo, so that to change pfrac (fraction of dynamo in solution in the NH. =1 low lat, =0 hi lat). The runs for two storms (20230420 & 20031120) has been carried out, the results are encouraging and as expected with comparison with the control runs by the WAM-IPE tip version. — Reply to this email directly, view it on GitHub <#89>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AH5BFOGL6MR7P7PCUC4N7GDX4WWELANCNFSM6AAAAAA5LJLPDY> . You are receiving this because you were assigned.Message ID: ***@***.***>

[astridmaute]

Hi Tim and Zhuxiao, I could not have guessed the magnitude of the TEC change, but Zhuxiao’s results make sense to me. It might be difficult to tune where the convection reversal boundary should be. We probably need to discuss how we pull together the different testing (crit value, SH convection, log interpolation …) to have an improved solution. At the last meeting you asked about the bias for the 20-21 Nov 2003 storm with linear and log10 ion density interpolation in the transport. On page 3-5 are some diagnostics<https://drive.google.com/file/d/1mvYhimJiOdnyWkMplk366gU7rQ7Cth9i/view?usp=drive_link> – page 3 is the mean of TEC compared to MIT TEC, page 4 is the mean of log10 TEC, page 5 shows both. I hope we can have a discussion once you are back. Enjoy the hiking, Astrid From: timfullerrowell ***@***.***> Sent: Thursday, September 28, 2023 1:29 PM To: NOAA-SWPC/GSMWAM-IPE ***@***.***> Cc: Astrid Maute ***@***.***>; Assign ***@***.***> Subject: Re: [NOAA-SWPC/GSMWAM-IPE] experiments on increase critical colatitude limits for use of Heelis potential in dynamo (Issue #89) Thanks Zhuxiao, Yes, these look promising, as you say. Be interesting to see the bias numbers for the recent storm against Glo-TEC. Tim

On Thu, Sep 28, 2023 at 11:15 AM ZhuxiaoLi ***@***.***<mailto:***@***.***>> wrote: Aiming to change the TEC distribution along low- midlatitude (make it more centralise toward equator) with changing the ion/electron circulation, Tim & Astrid designed a test to increase the critical colatitude limits for use of Heelis potential in dynamo. For the detailed IPE code, increase the values of offset1_deg, offset2_deg by 5 deg to change crit(1) & crit(2) in module_colath.f in ./dynamo, so that to change pfrac (fraction of dynamo in solution in the NH. =1 low lat, =0 hi lat). The runs for two storms (20230420 & 20031120) has been carried out, the results are encouraging and as expected with comparison with the control runs by the WAM-IPE tip version. — Reply to this email directly, view it on GitHub <#89>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AH5BFOGL6MR7P7PCUC4N7GDX4WWELANCNFSM6AAAAAA5LJLPDY> . You are receiving this because you were assigned.Message ID: ***@***.***<mailto:***@***.***>>

— Reply to this email directly, view it on GitHub<#89 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AFCVBFYDXKIM5SW6NGI2J73X4XFWVANCNFSM6AAAAAA5LJLPDY>. You are receiving this because you were assigned.Message ID: ***@***.******@***.***>>

[ZhuxiaoLi66]

Hi Tim,
What recent storm do you mean? I need the driver files after Aug.30 on Hera from Adam, already sent the request.
I bet this is an issue separate from the crit12 code? or you want to see the difference the code made for the bias number?
I sent you an email with the latest bias correction figure up to 20230728.

Thanks Zhuxiao, Yes, these look promising, as you say. Be interesting to see the bias numbers for the recent storm against Glo-TEC. Tim

[ZhuxiaoLi66]

Hi Astrid,
Thanks for sharing. I sent an access request for your shared link.
Agree the further discussion and please let me know how I can help more.
Zhuxiao

Hi Tim and Zhuxiao, I could not have guessed the magnitude of the TEC change, but Zhuxiao’s results make sense to me. It might be difficult to tune where the convection reversal boundary should be. We probably need to discuss how we pull together the different testing (crit value, SH convection, log interpolation …) to have an improved solution. At the last meeting you asked about the bias for the 20-21 Nov 2003 storm with linear and log10 ion density interpolation in the transport. On page 3-5 are some diagnosticshttps://drive.google.com/file/d/1mvYhimJiOdnyWkMplk366gU7rQ7Cth9i/view?usp=drive_link – page 3 is the mean of TEC compared to MIT TEC, page 4 is the mean of log10 TEC, page 5 shows both. I hope we can have a discussion once you are back. Enjoy the hiking,

[timfullerrowell]

The two storms that had the spikes in the bias (jumped from ~.9 to ~1.2) that prompted the study. I was just asking what that bias plot against Glo-TEC looks like after the change.

…

On Fri, Sep 29, 2023 at 1:52 PM ZhuxiaoLi ***@***.***> wrote: Hi Tim, What recent storm do you mean? I need the driver files after Aug.30 on Hera from Adam, already sent the request. I bet this is an issue separate from the crit12 code? or you want to see the difference the code made for the bias number? I sent you an email with the latest bias correction figure up to 20230728. Thanks Zhuxiao, Yes, these look promising, as you say. Be interesting to see the bias numbers for the recent storm against Glo-TEC. Tim — Reply to this email directly, view it on GitHub <#89 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AH5BFOF27BSCBWIOU63LELLX44RFPANCNFSM6AAAAAA5LJLPDY> . You are receiving this because you were assigned.Message ID: ***@***.***>

[ZhuxiaoLi66]

Got it, will do.

[ZhuxiaoLi66]

Hi All, the blue point in the following plot indicates the IPE TEC bias ratio against GloTEC on Apr.24 2023 after increasing the crit(1) & crit(2) by 5degs in the current experiment.
you can see the experiment brings the bias ratio down about 0.14 from 1.24 to 1.10, the magnitude is larger than the effect of the last experiment (1.19.vs.1.24, shown by the black point on the same day ) Astrid did to change fluxtube points at the equator with too small ksi (volume) values (IPE_plasma_class.F90).

[ZhuxiaoLi66]

Will show the result for Mar.24 2023 storm later.

[astridmaute]

Hi Zhuxiao, Thanks for the plots. I naively assume the original goal with the keddy factors is to adjust the neutral density but it will also affect the ion composition and TEC. We probably should keep an eye on both when doing any testing. If you have time it would be interesting to test the log interpolation for some time periods (storm) where you already have some results (I do not think it matters what code version). There would be two things to focus on 1. the mean TEC variation (or bias compared to a simulation you have) and 2. the number of flux tubes with issues. The latter one you can get by searching (grep) for 'Conv issue in O+ H+' in all the fcst.* files for the simulation and pipe it into a word count. You might be familiar with that e.g. grep "Conv issue..." fcst.* | wc -l If you have the fcst files from your original simulation that would be great to compare the flux tube numbers "with issues" if no we can compare with what I have found Below I add some instruction what would have to be changed to test the log interpolation. If you have time we can also talk about it tomorrow. This is only in case you have no other tasks to work on. Best, Astrid

…

----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- For the log interpolation you have to change the code in only tow places. In IPE_Plasma_Class.F90 SUBROUTINE Cross_Flux_Tube_Transport In the Q-interpolation change after !GHGM - Below the bottom of the tube at the Northern end Change density(1:n_conv_spec,lpx,mpx) calculation to density(1:n_conv_spec,lpx,mpx) = log10(plasma % ion_densities_old(1:n_conv_spec,isouth,lp_t0(lpx),mp_t0(mpx)))*i_comp_weight(1) + log10(plasma % ion_densities_old(1:n_conv_spec,inorth,lp_t0(lpx),mp_t0(mpx)))*i_comp_weight(2) after the mp,lp interpolation is done i.e. e_temperature(lpx,mpx)*lp_comp_weight(lpx)*mp_comp_weight(mpx) ENDDO ENDDO! "add" ion_densities_int(1:n_conv_spec) = 10**ion_densities_int(1:n_conv_spec) From: ZhuxiaoLi ***@***.***> Sent: Monday, October 2, 2023 1:09 PM To: NOAA-SWPC/GSMWAM-IPE ***@***.***> Cc: Astrid Maute ***@***.***>; Assign ***@***.***> Subject: Re: [NOAA-SWPC/GSMWAM-IPE] experiments on increase critical colatitude limits for use of Heelis potential in dynamo (Issue #89) Hi All, the blue point in the following plot indicates the IPE TEC bias ratio against GloTEC on Apr.24 2023 after increasing the crit(1) & crit(2) by 5degs in the current experiment. you can see the experiment brings the bias ratio down about 0.14 from 1.24 to 1.10, the magnitude is larger than the effect of the last experiment (1.19.vs.1.24, shown by the black point on the same day ) Astrid did to change fluxtube points at the equator with too small ksi (volume) values (IPE_plasma_class.F90). [TEC_bias_ratio_20211101_20230429_para5_crit12_5deg_qksi]<https://user-images.githubusercontent.com/22546571/272063055-04b5d5a0-6146-4380-aaa6-8eb9598857ab.png> - Reply to this email directly, view it on GitHub<#89 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AFCVBF72FBZPEE5YCB5MP3LX5MGNBAVCNFSM6AAAAAA5LJLPD2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONBTGYYDINJTGU>. You are receiving this because you were assigned.Message ID: ***@***.******@***.***>>